1 s2.0 S0308521X22001561 Main

Agricultural Systems 203 (2022) 103520
Contents lists available at ScienceDirect
Agricultural Systems
journal homepage: www.elsevier.com/locate/agsy
New trends in the global digital transformation process of the agri-food

sector: An exploratory study based on Twitter
María Ancín a, *, Emilio Pindado b, Mercedes Sánchez a
a
Department of Business Management and Institute on Innovation and Sustainable Development in Food Chain - (ISFOOD), Public University of Navarre, Pamplona,
Spain
b
Department of Agricultural Economics, Statistic and Business Management, ETSIAAB, Technical University of Madrid, Madrid, Spain
H I G H L I G H T S G R A P H I C A L A B S T R A C T
• The adoption of new digital technolo

gies in the agri-food value chain has
been influenced by the COVID-19
pandemic crisis.
• Data mining of Twitter content is useful
to analyse the perceptions towards new
digital technologies in the agri-food
sector.
• About 80% of the tweets regarding dig
ital transformation of the agri-food
sector showed a connection to positive
emotions.
• Understanding of the new digital tech
nologies within the sector is related to
sustainability and climate change
concerns.
• USA, India, UK, and Nigeria show
higher number of tweets that relate new
digital technologies to the agri-food
sector.
A R T I C L E I N F O A B S T R A C T
Editor: Laurens Klerkx CONTEXT: The agri-food system is undergoing pervasive changes in business models, facilitated by the use of
digital technologies. Although today it is almost inevitable for any business to adopt some level of digital
Keywords: transformation to strengthen their competitiveness, this transition in the agri-food sector could be more complex,
Digital transformation given its characteristics.
Agri-food sector
OBJECTIVE: The aim of the study is to analyse worldwide the perceptions of new digital technologies in the agri-
Agriculture 4.0
food sector expressed within social media platforms, identifying the differences that may exist between them
Twitter
Data mining regarding its objectives and social acceptance.
METHODS: This paper examines the information regarding digital transformation process in the agri-food sector
disseminated worldwide on Twitter. For that purpose, Twitter API is used to gather tweets and descriptive and
content analyses, including a sentiment analysis, are performed using R and MAXQDA software.
RESULTS AND CONCLUSIONS: We found that the digitalization of the agri-food sector is broadly discussed
within Twitter. Different actors participate in these information flows, being companies and digital solution
* Corresponding author.
E-mail address: maria.ancin@unavarra.es (M. Ancín).
https://doi.org/10.1016/j.agsy.2022.103520
Received 11 April 2022; Received in revised form 14 September 2022; Accepted 16 September 2022
Available online 23 September 2022
0308-521X/© 2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
M. Ancín et al. Agricultural Systems 203 (2022) 103520
providers the most active users and academics and governmental institutions the most visible. Artificial Intel
ligence was the most mentioned technology, that together with the Internet of Things, Big Data, Machine
Learning, and Cloud Computing, was related to improving production efficiencies, crop yield, or cost reduction.
In the case of Blockchain Technology, it was closer to food supply chain actors, such as distribution companies
and marketers. However, all these technologies are connected to the concept of sustainability. The sentiment
analysis showed a generally positive tone, indicating social acceptance regarding the starting phase of the
adoption of these technologies. The study also identified differences among countries, pointing to a stronger level
of engagement with these technologies in developed regions. Moreover, the COVID-19 pandemic was seen as a
chance to boost the digital transformation in the sector all over the world.
SIGNIFICANCE: Our results demonstrate that data harvested from Twitter provide useful insight into perceptions
of digital transformation and different digital technologies in the agri-food value chain across different countries.
Information that could be useful for researchers, but also for agricultural firms and policymakers.
1. Introduction the diffusion and adoption of this digital transformation, which will
serve to define policy interventions promoting it (Parra-López et al.,
The process of innovation development and adoption is perceived as 2021; Rijswijk et al., 2021). To fill this gap, the present study aims to
complex, dynamic and uncertain (Silvestre and Ţîrcă, 2019). However, explore worldwide the public perceptions towards different new digital
it is considered essential to enhance firms' sustainable performance, to technologies in the agri-food sector and identify the putative differences
achieve the Sustainable Development Goals (Sachs et al., 2019). In this regarding its social acceptance. We address this objective using the
regard, sustainable innovation adopted by firms and the entire supply content published on this topic within a popular social media platform.
chain must include the adoption of new digital technologies aimed at In this regard, social media has become a usual tool to express
improving the health of the planet and people, as well as efficiency and opinions and sentiments or just share relevant information, changing the
firm performance (Mondejar et al., 2021). Additionally, innovation in way in which people communicate (Li and Kent, 2021). This phenom
the agri-food sector has also been described as imperative to respond to enon presents opportunities and challenges for companies that are
new challenges related to sustainability, changing demand, and increasingly adopting social media as a communication channel,
increased competition. The COVID-19 pandemic has also brought to changing the ways they operate and relate to stakeholders (Paniagua
light a need to build resilient food systems, motivating innovative col et al., 2017). For these reasons, social media data has become a popular
lective actions along the entire agri-food chain (Bakalis et al., 2020). information source for academic research and other actors. Although it
The use of new digital technologies, such as the Internet of Things has been mainly focused on the analysis of food consumer behaviour
(IoT), Big Data, Artificial Intelligence (AI), and Blockchain Technolo (Mishra and Singh, 2018; Mostafa, 2013; Pindado and Barrena, 2020),
gies, provides new opportunities to face these threats. Today, it is almost Twitter has been also used as an entry point for understanding percep
inevitable for any company to adopt some level of digital transformation tions of specific digital innovations in the agri-food sector by the “in
to strengthen its competitiveness (Lu, 2017; Verhoef et al., 2021). Dig dustry, farmers, and the broader public” (Duncan et al., 2021, p. 1185).
ital transformation is changing the way companies conduct business Among different social media platforms, Twitter was selected to
because it affects operational routines and creates new ways to network conduct this study because it is not only one of the most popular social
with customers, suppliers, and stakeholders (Cheng and Wang, 2021). It platforms, but it is also easy to find topics, trends, attitudes, and senti
is evident that the development of this new business model requires the ments on it (Chamlertwat et al., 2012). Through Twitter data analysis,
strengthening and redefining of firms' capabilities (Matarazzo et al., the main objective of this research is to assess the digital technology
2021; Tortora et al., 2021). However, the transition towards digitali development stage in the agri-food sector, identifying differences be
zation could be complex in the agri-food sector, characterised by small tween different technologies regarding its social acceptance across
and medium-sized enterprises (SMEs), and problems of generational different countries. We analyse and compare the level of engagement
renewal (Żmija et al., 2020), as well as a supposedly low level of In towards different types of digital technologies, given their different
formation and Communication Technology (ICT) skills and engagement technological intelligence functions and their disruptive character (Yang
(Marshall et al., 2020). et al., 2021a). The research technique starts with Twitter data collection
Some authors have addressed the broad topic of digital trans performed through the Twitter API (Academic Research access level).
formation in agriculture, but most of them are focused on technical as Keywords such as “digital transformation”, “artificial intelligence” or
pects of specific productions, such as improving productivity or logistic “agri-food” were used in the search query. The analysis of Twitter data to
processes (Hunt and Daughtry, 2018; Rutten et al., 2013; Wathes et al., extract intelligence combined descriptive and content analysis, consid
2008; Wolfert et al., 2017). The social science field has recently started ering frequencies and the study of sentiments and emotions. The find
investigating various aspects of digital agriculture in relation to farm ings of this study will help us to acquire insights into the current and
production systems, value chains, and food systems as a whole, future digitalization trends in the agri-food sector and calibrate the in
regarding technology adoption and adaptation, farmers' skills and atti fluence of social networks' impacts. In this regard, it will be possible to
tudes, or policy processes (Klerkx et al., 2019). Although there is data extract suggestions in order to improve the communication strategy
available regarding the innovation management of companies, there are regarding innovation adoption by different actors of the agricultural
few studies regarding digital transformation processes in this sector, and sector.
most of them are focused on the adoption of particular technologies by The paper is organised as follows. Section 2 presents a literature
farmers (Haberli et al., 2019; Yoon et al., 2020; Zeng et al., 2021). review of research work done in the use of social media data in the agri-
Furthermore, the recent COVID-19 pandemic has forced companies, food sector and offers an overview of the digital transition of the sector.
including the rural actors, to look into digital solutions to continue Section 3 shows the research procedure and methodology, and Section 4
functioning, but our knowledge about its effect on the sector is still presents the results. Section 5 presents the discussion and conclusions,
limited (Galanakis et al., 2021). including managerial and research implications and future research
Hence, a wider perspective including the overall societal acceptance perspectives.
of the digital transformation process of the agri-food sector is required to
understand the adoption level, the importance and dynamics of different
technologies, as well as the role played by different agents involved in
2
2. Background ethical implications (Eastwood et al., 2019) and the possible disagree
ment with agroecological approaches to reach a responsible innovation
2.1. Digital transformation of the agri-food sector in the sector (Rotz et al., 2019). Moreover, there is a concern about the
inequity in the design of digital farming innovations (Bronson, 2019)
Innovation in our society is considered necessary to solve the current and the unequal relationships of power that the digital revolution could
global challenges. In that regard, it has been argued that innovation promote between players in the food system (Bronson and Knezevic,
systems need to be mission-oriented, trying to focus research and in 2016), or even its potential to directly include and exclude these players
vestments on solving critical problems, providing a solution or a con from the generated benefits (Klerkx and Rose, 2020). On the other hand,
crete approach (Mazzucato et al., 2020). This view has also reached it has been reported that digital technologies could accelerate post-
agricultural innovation systems, enabling a food system transformation COVID-19 recovery (Rowan and Galanakis, 2020). In fact, the last
(Klerkx and Begemann, 2020; Klerkx and Rose, 2020). Among other report from the (World Bank, 2021) outlines how digital technologies
concepts, digital technology implementation and development (i.e., based on data are springing up in transforming sectors, among which we
digital transformation) is one of the main pillars supporting the can find agriculture.
achievement of these challenges in the agri-food system (Shepherd et al., However, evolving from digitization to digital transformation ap
2020). This kind of technology can be also considered as part of the Key pears to be challenging in the sector. Some reviews have identified the
Enabling Technologies (European Commission, 2009), which are seen as most common barriers to the adoption of digital technologies in the agri-
drivers for innovation applicable in multiple industries. food sector (Annosi et al., 2020; Giua et al., 2020; Jespersen et al.,
The digital transformation process is a complex phenomenon dis 2013). Some of them are related to firms' internal resources, such as
cussed by experts, researchers, policymakers, and entrepreneurs, and it farm business characteristics, size, infrastructure, and financial avail
is considered to be a radical change in economy and production on a ability (Bronson, 2019; Lawson et al., 2011). Other challenges are
global scale (Kaplan and Haenlein, 2019; Verhoef et al., 2021). In the related to external resources, data complexity, transfer, and privacy,
recent academic literature it is possible to identify three phases in the internet connectivity, the lack of appropriate incentives, and a suitable
digital transformation process: digitization, digitalization, and digital adequate legislative operating environment (Kernecker et al., 2020;
transformation (Brenner and Hartl, 2021; Verhoef et al., 2021). The Pivoto et al., 2019). Regarding firm factors, the development of specific
concept of digitization refers to the conversion of information into competencies, named dynamic capabilities, is also considered essential
digital formats by computers, in a bid to enhance efficiency (Loebbecke to achieve the digital transformation of businesses (Matarazzo et al.,
and Picot, 2015). Digitalization entails a deeper transformation that 2021; Warner and Wäger, 2019). Finally, factors related to farmers'
changes value creation activities or existing business processes, such as personal characteristics, such as age, education, skills and knowledge of
enhancing customer experiences, communication and distribution using ICT, perceived profitability, environmental-related behaviours,
(Leviäkangas, 2016; Ramaswamy and Ozcan, 2016). Finally, digital availability of time, or simply the willingness to implement new tech
transformation is the most pervasive phase, that involves the emergence nologies, could be decisive to make possible the adoption of new digital
of entirely new business models considered new to the focal firm or technologies in the farming and food sectors, allowing its digital trans
industry (Brenner and Hartl, 2021; Verhoef et al., 2021). It can be also formation (Alvarez and Nuthall, 2006; Bowen and Morris, 2019; Fountas
argued that the digital transformation process could be different et al., 2015; Tey and Brindal, 2012).
depending on the size and sector of the companies concerned because it All of these emergent digital technologies have been considered as
is determined by some important characteristics, such as knowledge, part of the game-changing innovations that will transform food pro
R&D intensity, and technological assets (Aboelmaged, 2014; Chatterjee duction (Klerkx and Rose, 2020). However, given the aforementioned
et al., 2021). In relation to these assets, (Garzoni et al., 2020) introduced barriers to technology adoption, it is reasonable to assume that Agri
a four-level approach to SMEs' engagement in the adoption of digital culture 4.0 is still limited to a few innovative firms (Zambon et al.,
technologies: digital awareness, enquirement, collaboration, and trans 2019), indicating that the sector could still be in a starting phase inside
formation. This outlook could be even more complicated when the di the digital transformation process. Currently, there is a lack of aware
versity and complexity of the new digital technologies are taken into ness about which are the most frequently adopted technologies, their
account (Ciarli et al., 2021). These can be categorised into four groups limitations, and the conception of the agents involved, as well as those of
regarding their main application: efficiency technologies (e.g., Cloud society, all of which could strongly influence this process. More in
Computing), connectivity technologies (e.g., IoT), trust disintermedia vestigations are needed to explore if the theory around new digital
tion technologies (e.g., Blockchain), and automation technologies (e.g., technologies has been put into practice.
AI and Big Data) (Brenner and Hartl, 2021; Lanzolla et al., 2020).
The implementation of these technologies in the agri-food sector has 2.2. Agri-food sector and social media data
been happening for decades, however, it was mainly considered behind
the concepts of precision agriculture, smart farming, digital agriculture, Nowadays, social media applications, such as Facebook, LinkedIn,
agriculture 4.0 or farm management decision systems (Klerkx et al., Twitter and YouTube, among others, are widely used in diverse settings
2019). All of these terms could be part of the digital transformation and with different purposes. As an increasing number of users are
because implies that management tasks in any part of the food system moving towards social media platforms, companies find it imperative to
are based on data obtained from the use of different technologies use social media for brand building in order to create opportunities for
(Duncan et al., 2021; Eastwood et al., 2019). However, Wolfert et al., customer engagement (Schaefers et al., 2021; Shawky et al., 2020). Over
(2017) discern between precision agriculture and smart farming, the past decade, social media has been recognized as a key strategic
considering that the former is only focused on in-field variability but element of companies' competitiveness (Braojos et al., 2019). For agri-
does not takes into account data. This idea could be refutable, given that food firms, the usage of social media has grown very fast. Indeed, the
precision agriculture involves a range of technologies that generate data food industry is at the forefront of innovation in interactive marketing
to help in decision making (Bronson and Knezevic, 2016; Duncan et al., (Caiazza and Bigliardi, 2020). As an example, some of the major brands,
2021). Nowadays, it seems that these enabling technologies have such as Dr. Pepper, Kellogg's and CocaCola, have experienced increased
reached the whole food supply chain as part of the so-called Industry 4.0 attention from digital marketing within these platforms (Montgomery
(Lezoche et al., 2020; Trivelli et al., 2019). et al., 2011). However, other studies indicate that agri-food SMEs and
On the one hand, although positive effects of this revolution in the cooperatives are still on the way to achieving effective communication
agri-food system in terms of the increased efficiency, productivity, and and interaction with their target public in the digital environment
sustainability could be expected, there is a discussion about its socio- (Cristobal-Fransi et al., 2020).
3
Leaving commercial purposes aside, many issues related to the agri- First, data is stored as separate JSON files (JavaScript Object Notation),
food sector are likely to be present in social media (Stevens et al., 2016). but after applying a parsing method the output data can be stored as
Discussion and interaction of people regarding topics such as animal Comma Separated Values (CSV) or Excel files.
welfare, GMO and food safety are very frequent (Price, 2021). This Twitter data gathering can be difficult, considering that about 500
generates an opportunity for different actors in the agri-food system. For million tweets are generated per day (Karami et al., 2021). That makes
instance, social media applications are used for learning and collabo 6000 tweets every second, and that makes essential a data sampling
ration among experts, entrepreneurs, farmers (Chowdhury and Odame, process, based on keywords and hashtags, as well as a time period re
2013; Mills et al., 2019; Phillips et al., 2018), and even the education striction to extract relevant information. In this way, Twitter data
community (Aguilar-Gallegos et al., 2021). It provides the opportunity related to new digital technologies in the agri-food sector were collected
to overcome the physical distance between actors and create networks for a period of six months, from October 2020 to March 2021. This
directed to supporting agricultural innovation (Fielke et al., 2020). period of time, furthermore, was selected as representative of COVID-19
Hence, the analysis of social media data allows for studying real pandemic disruption in the agri-food sector which seems to have boos
events, social interactions, network analysis and user behaviour. In this ted the adoption of new digital technologies in the food industry and
regard, Twitter can be considered one of the most popular micro agriculture (Di Vaio et al., 2020; Rowan and Galanakis, 2020).
blogging platforms with 199 million monetizable daily active users and The data collection methodology consisted of retrieving tweets
500 million messages tweeted every day (Aslam, 2022), providing the containing any of these English words: “digitalization”, “digital trans
opportunity to interact with the audience without restriction and formation”, “big data”, “blockchain”, “artificial intelligence”, “AI”, “internet
disseminate information rapidly (Moe and Schweidel, 2017). Moreover, of things”, “IoT”, “machine learning”, “smart technologies”, “cloud
Twitter data is considered to be “open data” because it allows busi computing”, “smart agriculture”, “agriculture 4.0”, “smart farming”, “digital
nesses, practitioners or researchers to collect and analyse tweets through agriculture”, always accompanied by “agri-food”, “agrifood” or “agricul
its API. Consequently, it has been already used for academic research in ture”. The use of these terms as keywords in the search query was based
many fields (Karami et al., 2020) in order to analyse the current scenario on their popularity and their presence in previous works related to
but also to forecast upcoming trends related to any phenomenon in digital transformation trends in the agri-food sector and the adoption of
different locations. However, the study of Twitter to analyse the digital agriculture 4.0 or smart farming (Annosi et al., 2020; Klerkx et al., 2019;
transformation process in general, and especially in the agri-food sector, Rose and Chilvers, 2018)2. In this way, a majority of messages related to
is at a very early stage. Only a few authors have used social media data to the broad topic of digital innovation in the agri-food sector were
analyse attitudes towards new digital technologies without considering captured. An initial sample of 27,787 tweets was1 obtained, but due to
the agri-food sector, such as public perception of the IoT (Bian et al., the possibility of redundancies, duplicate tweets were removed, leaving
2016) or Blockchain Technology (Mnif et al., 2021), the intention to 27,500. A filter criterion was applied and only geolocalised tweets were
accept robotics in the workplace (Sinha et al., 2020) and AI utilisation considered, obtaining a final number of 18,001 tweets that composed
(Grover et al., 2020), in some cases combined with other methodologies. the working dataset.
For that reason, our study tries to specifically analyse the perception of Although the above methodology to extract information from
these technologies in the sector. Another important aspect is the possi Twitter has been extensively used to investigate trends and public
bility to analyse sentiments to gain an understanding of the opinions opinions regarding specific topics such as the digital transformation of
from which the engagement could be addressed (Caetano et al., 2018; tertiary industries (Sullivan et al., 2021), it is important to point out its
Veltri and Atanasova, 2015). The emotional component of tweets has potential to properly contextualize this research. Mainly, it cannot be
been used to understand social perception towards a specific phenom considered that social media users are representative of the general
enon also in the agri-food sector, such as precision agriculture (Ofori and population due to their users tend to be younger and more educated than
El-Gayar, 2021), consumer opinions towards food attributes (Borrero non-users (Vaccari et al., 2013). Consequently, research using this
and Zabalo, 2021; Samoggia et al., 2020) and new food trends (Pindado source of information should be interpreted according to this self-
and Barrena, 2020). However, these studies do not go in depth in the selection (Mellon and Prosser, 2017). Furthermore, there may be other
different technologies or the user profile that are key aspects of the potential biases like the organic nature of the data —instead designed
current study. data—, bot intervention, dependence on the quality of the search terms,
or the “the black box of APIs” (Chen et al., 2022). However, this data is
3. Research procedure especially useful to study emerging research topics due to the newness of
the data, the inherent features of social media discussions (e.g., they are
3.1. Data collection emergent and actual), and the easiness to capture data over large time
periods (Chen and Tomblin, 2021; Groves, 2011; Klašnja et al., 2017).
In this research, we acquired Twitter data using R software and the
Twitter application programming interface (API) through the “aca 3.2. Analytical framework
demictwitteR” package (Gentry, 2015). Twitter API, through the Aca
demic Research access level, enables access to the full archive of tweets We have to consider that the captured data and metadata from
published on Twitter, providing a way to collect tweets and metadata. Twitter is in the form of unstructured text (informal expressions) but, at
To get this access several requirements must be complied with, and an the same time, it is enriched (hashtags, followers, users, etc.) compared
application form must be completed and approved1. To access the to traditional data stored in companies' databases. It makes the analysis
Twitter API the API key, API secret, access token and access token secret more challenging, especially without an available methodology,
are required. These can be obtained from the Twitter Developer Plat although some analytical frameworks have been proposed (Chae, 2015;
form. The “academictwitteR” package allows researchers to collect Mishra and Singh, 2018). In this way, it turns out to be necessary to
tweets containing specified words or sets of words, but it is possible to apply some research methods to mine intelligence from social media
specify more complex queries to incorporate into the API call, such as data. The detailed description of the proposed analytical framework that
geographic location, URLs and media content (Barrie and Ho, 2021). includes such methods is depicted in Fig. 1. It consists of two method
ologies: descriptive analysis (DA) and content analysis (CA), the latter
1
We would like to thank the Twitter developer community for permission to
2
access the Academic Research product track and gather a high number of We previously conducted a series of searches on Twitter using MaxQDA
tweets. Software, that led us to find out which were the most relevant words.
4
Fig. 1. Proposed Twitter analytics framework.
encompassing a sentiment analysis. Furthermore, the tweets were the number of occurrences of a word in a tweet or at the entire dataset
clustered by country (based on the most active countries in Twitter) or level. The hashtag analysis (frequency and association) served as an
divided into six major digital technologies. Concretely, we use a tech information source about the fields that were more related to the topic of
nology classification that allows to capture the main I4.0 technologies interest and demonstrated how popular the topics are and how topics are
within the agri-food sector discussed by Yadav et al. (2022) namely related (Petersen and Gerken, 2021).
“Blockchain”, “IoT”, “Big Data” and “Cloud Computing”. Furthermore, Sentiment analysis can be considered to be part of CA, dealing with
we also classify the tweets that address “Machine Learning” technologies people's opinions, attitudes, and emotions about any topic expressed in
due to their relevance to the agricultural sector, as well as “AI” tech written texts (Liu, 2015). In general, sentiment analysis methods are
nologies due to this term tend to be used interchangeably with machine based on opinion extraction and sentiment classification into positive,
learning despite it is a specific approach of “AI” (Storm et al., 2020). negative, or neutral categories. The analysis can be performed on an
entire tweet dataset to reveal the overall sentiment, but it can be also
3.2.1. Descriptive analysis (DA) applied to clusters or specific sub-groups. One of the most widespread
The main social analytics techniques include a descriptive analysis methods is called the dictionary-based method, based on using a pre-
that allows for establishing the nature of both tweets and users when it is existing lexicon with information about which words and sentences
applied to analyse information coming from Twitter (tweets and meta are positive and which are negative (Wilson et al., 2009) so that senti
data). It is based on various metrics and statistics, such as the number of ment scores are calculated by pointwise mutual information measures.
tweets, the number of hashtags, unique users, retweets and classification We followed this method using the Syuzhet dictionary (Jokers, 2017),
of tweets into different types, among others (Bruns and Stieglitz, 2013). using the “Sentimentr” package (Rinker, 2017), which takes into ac
This analysis is especially useful for intelligence extraction in Twitter count contextual valence shifters of the sentences contained in each
analyses, given the enormous size and the enriched nature (e.g., users, tweet. Finally, we conducted an analysis of variance (ANOVA) and
hashtags, and URLs) of Twitter data. A broad but crucial view of the data Tukey's HSD post hoc test to address significant differences among the
could be obtained as prior knowledge on which to base a more detailed average emotion scores of each digital technology.
analysis. The information regarding the number of tweets per user, re
plies or retweets of each tweet shows the most active or visible users, 4. Results
groups of users regarding their activities, and other useful user-related
information (Joseph et al., 2017). It must be recalled that other met 4.1. Descriptive analysis (DA)
rics could be used to address different problems but they must be
appropriately selected for each case of study. Tweet metrics. The collecting method described in the previous sec
tion resulted in the identification of 18,001 original tweets involved in
3.2.2. Content analysis (CA) the use of digital technologies in the agri-food sector (without consid
CA allows for mining intelligence from the captured tweets that are ering retweets). This highlights the relevance of Twitter as a way to
in the form of unstructured texts (Krippendorff, 2004). An important spread the latest trends about digitalization in the sector. Although the
step in text analysis and classification, previous to the CA, is data pre- number of tweets could be considered stable over the analysed period,
processing (e.g., cleaning, removing noise), which includes the we can see a decrease in the number of those posted during the Christ
removal of links, non-Latin characters, numbers and users (Pindado and mas season and also a periodic decrease at weekends (Fig. 2). 45% of
Barrena, 2020). CA can be performed by automatic text processing tweets in this data collection were retweeted by other users and close to
methods, based on text capturing and machine learning algorithms. 14% of them were replied to. Additionally, 38% of tweets refer to other
First, text capturing techniques (Weiss et al., 2005) transformed the users and address messages to them. We found tweets coming from a
initial unstructured text into formatted data. Then, the techniques that total of 142 countries. However, there were important differences
we used for mining intelligence were word frequency analysis, term among them regarding the number of posted tweets due to the use of
association analysis, thematic clustering, hashtag analysis and sentiment English words in the search query language. The USA was the leading
analysis (Mishra and Singh, 2018). Word frequency is based on detecting country in terms of the number of tweets with 28.6% of the total dataset.
5
Fig. 2. Tweets posted during the analysed period.
It was followed by India, the UK and Nigeria, with means of 14.3%,

Table 2
10.9%, and 5.6%, respectively. The 25 most active countries represent
Most visible users (top 10 users by number of retweets and replies).
>90% of tweets in the dataset (Supplementary Table 1).
Regarding hashtags (i.e., words or phrases prefixed by #) that User name N◦ retweets + replies Location Profile
highlight certain topics, we found 7970 different hashtags in the tweets STPI 3988 India Government
with 65% of them (11,649 tweets) containing one or more hashtags. Dr.Omkar Rai 2411 India Individual
Iain Brown, PhD 1680 UK Individual
Additionally, nearly 90% (16,102) of the total tweets contained one or
World Economic Forum 1522 Switzerland Group
more URLs. The most popular URLs were companies' websites and ar NITDA Nigeria 754 Nigeria Government
ticles about initiatives or examples that described how to implement the nelson chamisa 719 Zimbabwe Government
new digital technologies in agriculture. We also clustered the tweets in akin alabi® 619 Nigeria Individual
terms of six popular digital technologies found in the dataset. From 蔡英文 Tsai Ing-wen 578 Taiwan Government
Thabi Leoka 557 South Africa Individual
18,001 tweets, 4125 specifically mention AI technology (23% of tweets).
yadu yadav 553 India Individual
It was followed by blockchain and IoT with 7.3% and 6.7%, respectively.
Machine Learning and Big Data are less frequently mentioned and are
present in 4.5% and 3.6% of the tweets, respectively. Cloud computing is visible ones. The company FarmWise was one of the most active ac
noticeably the least popular technology (0.3%). It is worth mentioning counts and it was also in a good position regarding its visibility (20th
that half of the tweets did not contain a specific mention of any of these position). The most active users tend to be companies, mainly from
technologies, they were related to more general concepts such as digi North America, India, or Europe (Table 1). In the case of the most visible
talization or smart farming. users, they are mainly individual accounts, some of them belonging to
User metrics. We found 9004 unique users in the dataset. It means governmental institutions, predominantly from Asia and Africa
that each user posted two tweets on average. However, this does not (Table 2). Moreover, we found users with different profiles, such as
reflect reality, because we found important differences regarding the consultants, media, or foundations/groups, pointing to different tech
activity of users. Indeed, we found that 10% of users accounted for 47% nology adoption phases depending on these more dynamic agents. We
of tweets. The most active users were calculated based on the number of observed that 7% of unique users identified were verified accounts,
posted tweets (Table 1). At the same time, we have the most visible which are “accounts of public interest”, corresponding to significant en
users, defined by a higher number of received retweets and replies tities such as academic institutions, governments, politicians, news or
(Table 2). Comparing both kinds of users, we found that they were not ganizations, journalists, companies, activists, as well as other influential
the same; the users with more tweets were not necessarily the most individuals3.
Regarding the number of followers —the most basic popularity
measure of Twitter users—, we found that >70% of the users had be
Table 1 tween 100 and 10,000 followers (Fig. 3), indicating that the majority of
Most active users (top 10 users by number of tweets). users interested in the digital transformation of the agri-food sector
User name N◦ tweets % Location Profile cannot be considered to be opinion leaders. However, we found 12 users
Future of Ag 759 4.22 USA Individual
with >5 million followers (Fig. 3), mainly accounts of media firms or
Amolexis Ltd 214 1.19 UK Company news agencies. It is worth mentioning that the account of the United
ukiot.store 153 0.85 UK Company Nations is involved in this topic, as also is Microsoft.
Suriya Subramanian 146 0.81 UK Consultant
FarmWise 137 0.76 USA Company
agriculturerobots 113 0.63 Germany Individual
Valuer News 95 0.53 Denmark Company
akin alabi® 93 0.52 Nigeria Individual
UrbanVN 80 0.44 Canada Company 3
See https://help.twitter.com/en/managing-your-account/about-twitter-ve
FOUNDERS CUBE 75 0.42 India Company
rified-accounts
6
We further performed word analysis in the four most active coun

tries: USA, India, UK, and Nigeria. In this case, we kept the words that
are part of the search query in order to analyse the differences regarding
the technologies among countries. We showed that AI was the most
often mentioned technology in these four countries (Supplementary
Table 2 and 3). The least common technology was the same in all
countries; Cloud Computing. However, we found differences regarding
the popularity of the remaining technologies among countries. IoT and
Blockchain are the second and third most popular technologies in all of
them except for Nigeria, where Big Data is more frequent than these two
technologies. Regarding COVID-19, we found that around 3.2% of
tweets in both the USA and UK were associated with this pandemic. By
contrast, India and Nigeria only showed 1.8 and 1.3% of tweets directly
related to COVID-19. We further clustered the tweets in terms of six
popular digital technologies (Blockchain, IoT, Big Data, Cloud
Computing, AI and Machine Learning) and then conducted word anal
ysis (Table 4). When we analysed the association of each technology
with COVID-19, we found that Big Data, Blockchain and AI seems to be
more related to the pandemic (close to 3% of tweets) compared with the
remaining technologies (lower than 1% of tweets). In the case of Cloud
Fig. 3. Frequencies of users involved according to the number of followers. Computing, we could not find any reference to the COVID-19 pandemic.
Hashtag analysis. In our dataset 7971 hashtags were found and they
4.2. Content analysis (CA) appeared 53,766 times. Moreover, 65% (11,649) of these tweets con
tained one or more hashtags. The most popular hashtags, without
We performed an in-depth analysis tackling tweets' content through considering the digital technologies, were: #agriculture, #technology,
a word, hashtag, and sentiment analysis. Word analysis involved word #farming, #smart, #agtech, #agitech, #futureofag, #digital, #inno
frequency and word combination frequency at the dataset and clustered vation, #sustainability, #robotics, #digitaltransformation, #farmers,
document levels according to tweet location (countries) or six top #datascience, #agribusiness, #automation, #food, #startup and #cli
technologies (AI, Big Data, IoT, Blockchain, Machine Learning and matechange, among others. It has not escaped our notice that hashtags
Cloud Computing). Hashtag analyses included hashtag frequency anal related to the COVID-19 pandemic were also present, but to a lesser
ysis. Sentiment analysis, including polarity and emotion analysis, was extent (e.g., #covid19, #covid, #coronavirus, #pandemic).
conducted of the entire number of tweets and clustered documents (by Sentiment analysis. Fig. 4 shows the sentiments at the entire dataset
country or technology). level. The most important idea was that >80% of tweets (14,691) tended
Word analysis. The most popular words in tweets (once we remove to be slightly positive with a score from 0 to 1. Close to 10% were
those that were used as keywords in the search query) were food (found considered neutral (score = 0) and only 8% of tweets were slightly
in 1740 tweets), climate (1606), farmers (1526) and technology (1319), negative with a score from 0 to − 1. Very few tweets showed a relatively
among others (Table 3A). We then analysed how many times a particular strong positive or negative sentiment.
sequence of two words appears in the dataset (Table 3B). This provides Table 5 shows some exemplar tweets with a relatively positive,
information regarding those aspects which were getting attention on neutral or negative sentiment. When we performed this analysis in the
Twitter related to the digitalization of the agri-food sector. It is worth clustered documents by countries or digital technologies, we found a
mentioning that words related to the COVID-19 pandemic were present similar pattern to that of the entire dataset, but we could not find sig
in the data set, such as “covid”, “pandemic”, “vaccine”, “coronavirus” nificant differences among the countries or technologies for each cate
and “lockdown”. In general, 2.5% of tweets contained a reference gory of sentiments.
related to the COVID-19 pandemic, showing a link with the use of digital In order to go into detail about sentiment analysis, we performed an
technologies in the agri-food sector. Some of the most frequent word analysis of emotions, mining different kinds of sentiments from the
combinations were: “impact of covid”, “agriculture market covid” or tweets. In Fig. 5, we can see the most frequent emotions in the entire
“producers overcome covid”. dataset. The emotions considered to be positive were predominant,
especially trust and joy. When we analysed the emotions in the clustered
documents by technology, we found that they were quite similar to those
Table 3 observed for the entire dataset (Fig. 5). However, among technologies, it
Word analysis. A) Word frequency; B) Two-word combination frequency. can be highlighted that there was a significantly higher sentiment of
A) Word Freq % B) 2-word combination Freq % trust in tweets associated with machine learning compared with the
food 1740 9.67 climate smart 1250 6.94
other technologies. Another positive emotion, as is joy, was more
climate 1606 8.92 agriculture market 596 3.31 associated with machine learning and AI, followed by blockchain.
farmers 1526 8.48 precision agriculture 340 1.89 However, it can be seen that AI is also more significantly related to fear,
technology 1319 7.33 supply chain 265 1.47 which is a negative emotion, compared with other technologies.
new 1295 7.19 food security 256 1.42
market 1166 6.48 climate change 220 1.22
industry 1036 5.76 agriculture sector 225 1.25 5. Discussion and conclusions
help 942 5.23 agriculture industry 223 1.24
future 874 4.86 sustainable 175 0.97 The broad topic of digital transformation in the agri-food sector is
agriculture
addressed on Twitter because it was possible to identify >18,000 geo
agricultural 778 4.32 smart cities 156 0.87
global 765 4.25 food supply 153 0.85 localised related tweets in a period of 6 months. However, compared to
sector 706 3.92 agriculture platform 153 0.85 the 500 million tweets that are sent per day (Karami et al., 2021), this is
solutions 706 3.92 food production 140 0.78 a very small proportion. We found that the USA was the main country
farm 688 3.82 digital technology 127 0.71 involved in this topic, as well as the leading country in terms of the
world 640 3.56 food systems 123 0.68
number of Twitter users worldwide (Statista, 2021). However, we
7
Table 4
Detailed word analysis (2-word combination frequency) in clustered documents by technology.
Blockchain IoT Big data
2-word comb. Freq % 2-word comb. Freq % 2-word comb. Freq %
supply chain 182 13.8 smart agriculture 88 7.3 agriculture market 40 6.2
food supply 119 9.0 precision agriculture 88 7.3 artificial intelligence 40 6.2
global food 77 5.8 smart farming 61 5.1 the future 33 5.1
chain market 74 5.6 agriculture industry 50 4.2 future of 24 3.7
food security 73 5.5 real time 49 4.1 precision agriculture 23 3.6
digital agriculture 73 5.5 in 2021 45 3.8 agriculture industry 20 3.1
to track 68 5.1 to improve 45 3.8 platform for 20 3.1
agriculture giants 62 4.7 agriculture market 44 3.7 smart farming 18 2.8
track grains 62 4.7 crop yields 43 3.6 agriculture needs 16 2.5
ai strawberries 61 4.6 iot sensors 43 3.6 smart agriculture 15 2.3
agri food 50 3.8 potential to 42 3.5 can help 15 2.3
agriculture market 39 3.0 food production 41 3.4 cgiar platform 15 2.3
blue nova 39 3.0 machine learning 40 3.3 cruises over 15 2.3
supply chains 35 2.6 production costs 36 3.0 robotic buggy 15 2.3
covid 19 26 2.0 agricultural efficiencies 35 2.9 over crops 14 2.2
Cloud computing AI Machine learning
2-word comb. Freq % 2-word comb. Freq % 2-word comb. Freq %
grand farm 7 11.5 agriculture market 324 7.9 artificial intelligence 129 16.1
trilogy networks 7 11.5 can help 137 3.3 help to 55 6.8
precision agriculture 4 6.6 to improve 127 3.1 agriculture stimulates 53 6.6
rural cloud 4 6.6 the future 123 3.0 fresh produce 53 6.6
smart farming 4 6.6 machine learning 123 3.0 growth infrastructure 53 6.6
artificial intelligence 3 4.9 agriculture daily 116 2.8 need help 53 6.6
cloud based 3 4.9 future of 109 2.6 stimulates growth 53 6.6
farm launch 3 4.9 the potential 106 2.6 to improve 52 6.5
future of 3 4.9 potential to 103 2.5 end hunger 51 6.4
launch rural 3 4.9 in 2021 101 2.4 the potential 42 5.2
networks joins 3 4.9 precision agriculture 97 2.4 food production 41 5.1
provide cloud 3 4.9 improve agriculture 77 1.9 crop yields 40 5.0
retrieve from 3 4.9 smart agriculture 65 1.6 improve crop 40 5.0
send to 3 4.9 the world 64 1.6 in 2021 40 5.0
store in 3 4.9 agriculture industry 62 1.5 potential to 40 5.0
Table 5
Exemplar tweets with relative positive (score > 0), neutral (score = 0) or
negative (score < 0) sentiment.
Exemplar tweets Score
Advanced farming solutions to improve productivity, efficiency and 2.03

sustainability using deep learning https://t.co/jq4OQ7kI9T #agtech
#Augmenta #agriculture #automation #IoT #cloud
#ArtificialIntelligence #DeepLearning https://t.co/ROebs0c8wW
Digital farming is providing farmers new ways to provide more food with 1.80
more precise information. Learn how technology is changing farming for
the better: https://t.co/6ASBMZh0pY
Technology is of extreme importance in agriculture and #Plant_Scope helps 1.68
farmers find more efficient ways to protect their crops from diseases by
leveraging computer vision, Machine learning and Deep Learning to
monitor and precisely detect plant diseases. https://t.co/B2XlMbLoWT
Agricultural sensors for monitoring soil water and climate https://t.co/l 0
U8V8N6SEC
Viewpoint article from myself and @d_christianrose in @GeneticLiteracy: − 0.86
Genetics and AI have launched an agricultural revolution but ‘blind
techno-optimism’ could have harmful consequences. #Agriculture http
Fig. 4. Sentiment analysis at the entire dataset level. Polarity of tweets (− 3 s://t.co/fRT5Vm5ygc
more negative to +3 more positive). Obstacles to big data in agriculture: data error, inaccessible or unusable − 1.02
data, incompatible systems, inconvenience, unclear ROU and unclear
ownership #bigagdata https://t.co/DFunhpgnlJ
showed that the countries where Twitter is particularly popular (Sta
tista, 2021) are not necessarily the countries that were tweeting more
about digital technologies in the agri-food sector. In general, our data engaging with business partners or clients (Juntunen et al., 2020).
indicate that there were few countries with a relatively important ac The most visible users were individual users, some of them related to
tivity in this regard: USA, India, and UK. The analysis also revealed in academia, but also governmental institutions and politicians. It means
formation about the characteristics of the users. We found that some of that the information tweeted by this kind of user was more widespread
the most active users (higher number of tweets in the analysed period) than those posted by companies. In fact, it is said that a user retweets
were shown to be companies advertising digital solutions. However, when he has trust in the author (Firdaus et al., 2018), so it seems that
these companies did not have a high level of visibility (in terms of replies official or institutional Twitter accounts are more reliable in the view of
and retweets), indicating that the aim of their social media strategy is other Twitter users. Indeed, we found that percentage of verified
more related to promotional and advertising activities, rather than to
8
Fig. 5. Emotion analysis. Count of emotions indicates the strength of emotion present in the dataset (sum of each tweet). Emotions scores represent the average value
for each cluster of tweets by technology (mean ± SE). Different letters indicate significant differences (p < 0.05, ANOVA).
accounts (7%) in the dataset is higher than the average of Twitter 2%, et al., 2010; Chae, 2015), indicating that the content regarding digital
which reveals that verified accounts with supposed high-credibility play transformation process in the agri-food sector is more widespread.
an important role in the dissemination of the information regarding the Likewise, we found that the percentage of tweets that contained, at least,
digital transformation of the agri-food sector compared to other public one hashtag is higher than other studies analysing new technologies in
discussions (Yang et al., 2021b). However, despite the growing interest other sectors (Bougie et al., 2011). All these aspects indicate that tweets
in digital agriculture within policy circles (Klerkx et al., 2019), related to the digital transformation of the agri-food sector seem to be
governmental accounts do not seem to participate actively in the dis more conversational and engaging than random tweet samples or tweets
cussion of this process within the social media platforms. This is in line related to other topics. This finding reveals how Twitter may serve as a
with the fact that agricultural extension services traditionally have used useful platform for collective and individual learning regarding the
farmer field days and workshops, as well as face-to-face programs, to digitalization of this sector (Klerkx, 2021; Phillips et al., 2021).
diffuse agricultural innovations (Norton and Alwang, 2020). The use of Regarding the content, we found that more than half of tweets dealt
social media like Twitter has started to gain increasing relevance as a with general aspects of the digital transformation process in the sector,
diffusion mechanism for governments and agricultural extension ser such as smart farming or precision agriculture, rather than mention the
vices but they use these platforms to disseminate information in a top- use of specific digital technologies. However, it is interesting to specify
down approach with lower levels of engagement (Phillips et al., that a quarter of the tweets were related to AI, which seems to be the
2021). Consequently, we found that despite the higher levels of these most popular technology by far. This popularity could come from the
accounts in terms of their visibility they do not fully exploit Twitter as a conception of AI as an umbrella term that encompasses, in many cases,
platform of knowledge exchange for the digital transformation of the IoT (sensors that collect huge amounts of data), Big Data and Machine
sector (Klerkx, 2021). Learning (algorithms to analyse the data). Despite that, our results
The analysis also showed that users involved in the digital trans indicated increasing attention on the potential of AI in the agri-food
formation process of the agri-food sector could be less active compared sector. A recent study suggested a more active adoption of AI in North
to those interested in other topics in terms of the average number of America and Europe, although Asia and Africa were also making smaller
original tweets per user (Chae, 2015). Furthermore, the mentions of this but increasing efforts (Lakshmi and Corbett, 2020), that are in accor
topic on Twitter are not concentrated in a few users (Chae, 2015). These dance with our results. It is also remarkable that Big Data technology,
results, linked to the different profiles of these users (e.g., individual which is considered to be at an early stage but with high potential in
users, companies, institutions, media), indicate that a broad range of agriculture (Moysiadis et al., 2021), is one of the least mentioned digital
users were generating content on Twitter without clear leadership. This technologies, together with Cloud Computing. These results contrast
could suggest that different actors in the agri-food sector are involved in with the Future of Jobs Survey conducted by the (World Economic
the process of digital transformation, from farmers, producers, through Forum, 2020), where business leaders identify the most popular tech
to the food industry, the supply chain and finishing in the market. This nologies that are likely to be adopted by companies in the agriculture,
evidence is supported in the literature, where it can be seen that all agri- food and beverages sector in 2025: IoT and Big Data, followed by Cloud
food related stakeholders are making efforts to apply these technologies Computing and AI. It seems that AI catches more of the attention of the
that play a key role in their operations and decision-making (Lezoche general public compared with that of entrepreneurs in the sector and the
et al., 2020). reason could be that AI for the extended agri-food supply chain is only
Additionally, it must be considered that Twitter supports a variety of beginning to emerge (Monteiro and Barata, 2021).
communicative practices, and tweets are disseminated to converse with The word combination analysis of clustered documents by technol
individuals, groups and the general public. We found that the number of ogies allowed us to differentiate the activities of the agri-food sector that
direct conversations regarding digital transformation in the agri-food were prone to adopt each digital technology. All of them, except for
sector is higher than other studies addressing other topics within the Blockchain, were related to the digitalization of the agricultural pro
sector, such as supply chains (Chae, 2015). Moreover, the proportion of duction and highlighted the smart farming or precision agriculture
tweets that were retweeted is high compared to other studies (Boyd concepts within this digitalization process. This idea is in accordance
9
with a growing body of literature that recognizes the essential role of sector of developing countries to overcome the consequences of the
digital emerging technologies in precision agriculture (Charania and Li, pandemic.
2020) that allow for the developing of Decision Support Systems based Finally, trying to analyse the perception of new digital technologies
on data analysis and data mining. This links with the recurrent idea in the agri-food sector, we showed that the dataset contained a relatively
found in these tweets about the potential of these technologies to low sentiment level. This finding is not surprising given the content of
improve efficiencies, crop yields and reduce production costs, especially tweets is mostly focused on events, news or advertising, that differs from
in the case of AI and IoT. However, it seems that Blockchain technology the kind of tweets that usually have stronger sentiments, such as com
is mainly applied to distribution or commercialization because those plaints or discussions related to consumer behaviour (Pindado and
tweets are more related to the supply chain, traceability, and food safety. Barrena, 2020). Although we found a weak level of sentiment, we did
This evidence is supported in the literature, where it is possible to find a trend towards positivity, which is interesting considering the
identify Blockchain applicability to improve food quality, safety stan disruptive character of these technologies. In fact, the public may
dards and supply chain monitoring and tracking, especially when it is perceive digital technologies as a threat in the context of improving
integrated with IoT technology (Dey and Shekhawat, 2021; Kamilaris agricultural efficiencies (Driessen and Heutinck, 2015). However, our
et al., 2019; Torky and Hassanein, 2020). When the word combination results are in line with a study performed in Germany, where people
frequencies were analysed in the most active countries, what was said in showed a predominantly positive attitude towards the use of digital
USA, India and UK could be quite similar, focused on the use of these farming technologies (Pfeiffer et al., 2021). The analysis of specific
technologies mainly in agriculture. Although comparing India with USA emotions supports the idea of the positive tone in the data because trust
and UK could be noteworthy because this country is predominantly and joy showed higher scores compared to anger, fear, disgust, or
engaged in smallholder agriculture, India can be considered as an sadness. Most interestingly, it was possible to associate some of these
emerging economy focused on technology development and the view on emotions with specific digital technologies. We showed that AI-related
how to manage the digital transformation in the sector could be closer to tweets were significantly closer to the emotion of fear, which seems
those of developed countries (Mondal and Basu, 2009). However, the negative a priori. However, we realised that the emotion of fear was
speech in Nigeria was different, because most of the tweets were focused related to the concern about global issues, such as agriculture, food
on a specific initiative to promote climate smart agriculture in the demand or climate change, and AI was shown as a way to solve them.
country. Those differences can be considered understandable, taking The significantly higher sentiment of trust in tweets related to Machine
into account that the challenges facing the agri-food sector depend on Learning, but also to Blockchain and AI compared with the remaining
the economic status and development level of each country (Anasta technologies was also notable. The general trend in Twitter was to ex
siadis et al., 2018). As an example, AI technology seems to be unevenly press confidence towards the potential use of these new digital tech
distributed between developed and developing economies (Vinuesa nologies that could be indicative of a generalised acceptance by society,
et al., 2020). In general, it could be possible to understand that they are taking into account the early phase of adoption of these technologies in
in different phases of the digitalisation process, although the phenom the sector.
enon of digital transformation goes beyond the binary of developed and
developing countries (Freidberg, 2017), with an agricultural sector
habitually working in a global competitiveness context with a major 5.1. Implications
sustainability supply chain requirements.
The topics linked to the digital transformation process in the agri- Derived from our findings and the proposed approach it is possible to
food sector in Twitter were very diverse, evidenced by the high num draw some implications for government, rural actors, or researchers.
ber of hashtags in the data. Some of the most frequent words and First of all, we have to realise that the use of social media by the gov
hashtags in the dataset were related to concern about the environment, ernment has evolved from the distribution of propaganda to transparent
such as “climate change” or “sustainable agriculture”. This is not sur communication and engagement with the general public (Bonsón et al.,
prising, as it has been globally proclaimed that digitalisation is of critical 2019; Mergel and Bretschneider, 2013). Taking advantage of that new
importance to environmental sustainability (Wyckoff and Pilat, 2017). way of interaction, the methodology proposed here could be useful to
This is in line with the growing expectations about the potential envi identify attitudes towards governmental opinions, regulations, or sub
ronmental benefits of digital transformation in agriculture noted by the sidies regarding the digital transformation process in the agri-food
literature addressing both concepts (Del Río Castro et al., 2021; Isensee sector. Likewise, considering the findings reported in this study, public
et al., 2020). This view is close to the concept of S3enterprise that can institutions should improve their interaction with agricultural producers
define Agriculture 4.0: “sensing” (detect events, acquire data and mea and consumers within the social media platforms to improve the
sure changes that occur in a physical environment), “smart” (analyse knowledge exchange and learning mechanisms that may boost agricul
situations and make decisions based on the available data in a predictive tural digitalization (Klerkx, 2021). In the case of agricultural companies
or adaptive manner) and “sustainability” (optimise performance and farmers social media may be used to increase brand loyalty and
considering social, economic and environmental balance) (Miranda reputation (Swani et al., 2014). In that case, the descriptive analysis
et al., 2019). However, the empirical evidence that demonstrates the could indicate their popularity and reputation. Sentiment analysis of the
environmental gains as a result of the adoption of digital technologies in tweets mentioning them can enable the measurement of how the content
the agri-food sector is still scarce (Clapp and Ruder, 2020; Klerkx and of their tweets is perceived by professionals or clients. Additionally, this
Rose, 2020). approach allows them to identify what kind of users are reacting to their
Moreover, it was possible to extract an idea about the connection posts, and analyse if they are reaching the target audience or if they need
between COVID-19 and digital transformation in the sector because we to expand their brand community on social media sites (Zaglia, 2013).
found some tweets that regarded the COVID-19 pandemic as a digital Additionally, the research design shows the possibility of using Twitter
push (Amankwah-Amoah et al., 2021). It is true that the impacts of the data for research regarding the digital transformation process in the
COVID-19 pandemic and the response mechanisms have been different agri-food sector. Through the Twitter API, researchers have the oppor
for large companies, SMEs and small scale farming systems (Lopez- tunity to access Twitter data, which is interesting in terms of size, speed,
Ridaura et al., 2021). In this regard, we found that in the USA and UK and variety. The analysis of social media data has been recently
this idea is more broadly supported in social media than in other kinds of researched (Ghani et al., 2019), showing its potential as a new data
countries where the agriculture sector is in a different phase of devel source, which could complement the existing ones in the context of
opment. However, many tweets coming from Europe and the USA refer Social Representation Theory (Bäckström et al., 2003; Gaspar et al.,
to the potential of implementing digital technologies in the agri-food 2014; Ribeiro et al., 2016).
10
5.2. Limitations and future research org/10.1016/j.agsy.2022.103520.
Despite the contributions made, this research presents some limita References
tions. Starting from the data source, the use of Twitter data may not be
strongly demographically representative of the general population as Aboelmaged, M.G., 2014. Predicting e-readiness at firm-level: an analysis of
technological, organizational and environmental (TOE) effects on e-maintenance
users tend to be younger, more educated and live in urban areas (Bian readiness in manufacturing firms. Int. J. Inf. Manag. 34, 639–651. https://doi.org/
et al., 2016; Vidal et al., 2015). However, social media data could be 10.1016/j.ijinfomgt.2014.05.002.
useful for researchers as long as their limitations are recognized (Chae, Aguilar-Gallegos, N., Klerkx, L., Romero-García, L.E., Martínez-González, E.G., Aguilar-
Ávila, J., 2021. Social network analysis of spreading and exchanging information on
2015; Pindado and Barrena, 2020). Related to the data collection, the twitter: the case of an agricultural research and education Centre in Mexico. J. Agric.
selected period, keywords, and language, despite being selected to Educ. Ext. https://doi.org/10.1080/1389224X.2021.1915829.
address the specific objective of this study, could be considered to be Alvarez, J., Nuthall, P., 2006. Adoption of computer based information systems. Comput.
Electron. Agric. 50, 48–60. https://doi.org/10.1016/j.compag.2005.08.013.
limitations. In this regard, future research using Twitter data should use Amankwah-Amoah, J., Khan, Z., Wood, G., Knight, G., 2021. COVID-19 and
extended periods to corroborate the findings revealed here. Moreover, digitalization: the great acceleration. J. Bus. Res. 136, 602–611. https://doi.org/
the keywords that we used in the search query were carefully selected 10.1016/j.jbusres.2021.08.011.
Anastasiadis, F., Tsolakis, N., Srai, J.S., 2018. Digital technologies towards resource
but could have been different or included a wider range of concepts that
efficiency in the agrifood sector: key challenges in developing countries.
could be contemplated in future studies. Linked to this aspect, we only Sustainability 10, 4850. https://doi.org/10.3390/su10124850.
considered tweets posted in English because it is the predominant lan Annosi, M.C., Brunetta, F., Capo, F., Heideveld, L., 2020. Digitalization in the agri-food
guage on Twitter but retrieving tweets in different languages could industry: the relationship between technology and sustainable development. Manag.
Decis. 58, 1737–1757. https://doi.org/10.1108/MD-09-2019-1328.
contribute to the understanding of differences among countries and re Aslam, S., 2022. Twitter by the Numbers: Stats, Demographics & Fun Facts [WWW
gions. Likewise, the detailed analysis of the geolocation of the tweets Document]. Omnicore. URL. https://www.omnicoreagency.com/twitter-statistics/
could provide relevant insights into how the social representations of (accessed 12.7.21).
Bäckström, A., Pirttilä-Backman, A.-M., Tuorila, H., 2003. Dimensions of novelty: a social
digital transformation are determined. representation approach to new foods. Appetite 40, 299–307. https://doi.org/
With this overview, we consider that future research related to the 10.1016/S0195-6663(03)00005-9.
transition towards digitalization in the agri-food sector should be Bakalis, S., Valdramidis, V.P., Argyropoulos, D., Ahrne, L., Chen, J., Cullen, P.J.,
Cummins, E., Datta, A.K., Emmanouilidis, C., Foster, T., Fryer, P.J., Gouseti, O.,
focused on the factors that have an influence on its feasible imple Hospido, A., Knoerzer, K., LeBail, A., Marangoni, A.G., Rao, P., Schlüter, O.K.,
mentation, such as knowledge or awareness of the technologies, their Taoukis, P., Xanthakis, E., Van Impe, J.F.M., 2020. Perspectives from CO+RE: how
usefulness and perceived adoption costs, given that the disruptive digital COVID-19 changed our food systems and food security paradigms. Curr. Res. Food
Sci. 3, 166–172. https://doi.org/10.1016/j.crfs.2020.05.003.
technologies are in an emergent phase (Klerkx and Rose, 2020). Finally, Barrie, C., Ho, J., 2021. academictwitteR: an R package to access the twitter academic
we consider that the study of digital technology adoption from a social research product track v2 API endpoint. J. Open Source Softw. 6, 3272. https://doi.
perspective is urgently needed. The progressive adoption of digital org/10.21105/joss.03272.
Bian, J., Yoshigoe, K., Hicks, A., Yuan, J., He, Z., Xie, M., Guo, Y., Prosperi, M.,
technologies by firms implies the development of new skills and capa
Salloum, R., Modave, F., 2016. Mining Twitter to assess the public perception of the
bilities that will leverage each technology to the fullest to drive inno “Internet of Things”. PLoS One 11, e0158450. https://doi.org/10.1371/journal.
vation and optimise processes. In that regard, companies, including the pone.0158450.
rural alternatives, have to be aware of this phenomenon regarding Bonsón, E., Perea, D., Bednárová, M., 2019. Twitter as a tool for citizen engagement: an
empirical study of the Andalusian municipalities. Gov. Inf. Q. 36, 480–489. https://
digital transformation that accelerates the shift in workforce skills and doi.org/10.1016/j.giq.2019.03.001.
thus, the analysis of the whole range of factors that could act as drivers Borrero, J.D., Zabalo, A., 2021. Identification and analysis of strawberries’ consumer
and barriers to digital transformation would be required. opinions on twitter for marketing purposes. Agronomy 11, 1–19. https://doi.org/
10.3390/AGRONOMY11040809.
Bougie, G., Starke, J., Storey, M.A., German, D.M., 2011. Towards understanding twitter
CRediT authorship contribution statement use in software engineering: preliminary findings, ongoing challenges and future
questions. In: Proceedings of the 2nd International Workshop on Web 2.0 for
Software Engineering, pp. 31–36. https://doi.org/10.1145/1984701.1984707.
María Ancín: Methodology, Software, Validation, Investigation, Bowen, R., Morris, W., 2019. The digital divide: implications for agribusiness and
Data curation, Writing – original draft, Writing – review & editing. entrepreneurship. lessons from Wales. J. Rural. Stud. 72, 75–84. https://doi.org/
Emilio Pindado: Methodology, Data curation, Software, Supervision, 10.1016/j.jrurstud.2019.10.031.
Boyd, D., Golder, S., Lotan, G., 2010. Tweet, tweet, retweet: Conversational aspects of
Writing – review & editing. Mercedes Sánchez: Conceptualization, retweeting on twitter. In: 43rd Hawaii International Conference on System Sciences,
Supervision, Writing – review & editing, Funding acquisition. pp. 1–10. https://doi.org/10.1109/HICSS.2010.412.
Braojos, J., Benitez, J., Llorens, J., 2019. How do social commerce-IT capabilities
influence firm performance? Theory and empirical evidence. Inf. Manag. 56,
Declaration of Competing Interest 155–171. https://doi.org/10.1016/j.im.2018.04.006.
Brenner, B., Hartl, B., 2021. The perceived relationship between digitalization and
The authors declare that they have no competing interests (financial ecological, economic, and social sustainability. J. Clean. Prod. 315, 128128 https://
doi.org/10.1016/j.jclepro.2021.128128.
or personal relationships) that could have appeared to influence the
Bronson, K., 2019. Looking through a responsible innovation lens at uneven
work reported in this paper. engagements with digital farming. NJAS Wagen. J. Life Sci. 90–91, 1–6. https://doi.
org/10.1016/j.njas.2019.03.001.
Data availability Bronson, K., Knezevic, I., 2016. Big data in food and agriculture. Big Data Soc. 3 https://
doi.org/10.1177/2053951716648174, 205395171664817.
Bruns, A., Stieglitz, S., 2013. Towards more systematic twitter analysis: metrics for
Data will be made available on request. tweeting activities. Int. J. Soc. Res. Methodol. 16, 91–108. https://doi.org/10.1080/
13645579.2012.756095.
Caetano, J.A., Lima, H.S., Santos, M.F., Marques-Neto, H.T., 2018. Using sentiment
Acknowledgements analysis to define twitter political users’ classes and their homophily during the 2016
American presidential election. J. Intern. Serv. Appl. 9, 18. https://doi.org/10.1186/
This work has been supported by the Spanish Ministry of Science and s13174-018-0089-0.
Caiazza, R., Bigliardi, B., 2020. Web marketing in Agri-food industry: challenges and
Innovation, AEI/10.13039/501100011033 and the FEDER Funds (EU) opportunities. Trends Food Sci. Technol. 103, 12–19. https://doi.org/10.1016/j.
“Una manera de hacer Europa”, through the RTI2018-093791-B-C21 tifs.2020.04.024.
research project. Chae, B., 2015. Insights from hashtag #supplychain and twitter analytics: considering
twitter and twitter data for supply chain practice and research. Int. J. Prod. Econ.
165, 247–259. https://doi.org/10.1016/j.ijpe.2014.12.037.
Appendix A. Supplementary data Chamlertwat, W., Bhattarakosol, P., Rungkasiri, T., Haruechaiyasak, C., 2012.
Discovering consumer insight from twitter via sentiment analysis. J. Univ. Comput.
Sci. 18, 973–992. https://doi.org/10.3217/jucs-018-08-0973.
Supplementary data to this article can be found online at https://doi.
11
Charania, I., Li, X., 2020. Smart farming: Agriculture’s shift from a labor intensive to Groves, R.M., 2011. Three eras of survey research. Public Opin. Q. 75, 861–871. https://
technology native industry. Intern. Things 9, 100142. https://doi.org/10.1016/j. doi.org/10.1093/poq/nfr057.
iot.2019.100142. Haberli, C., Oliveira, T., Yanaze, M., 2019. The adoption stages (evaluation, adoption,
Chatterjee, S., Rana, N.P., Dwivedi, Y.K., Baabdullah, A.M., 2021. Understanding AI and routinisation) of ERP systems with business analytics functionality in the context
adoption in manufacturing and production firms using an integrated TAM-TOE of farms. Comput. Electron. Agric. 156, 334–348. https://doi.org/10.1016/j.
model. Technol. Forecast. Soc. Chang. 170, 120880 https://doi.org/10.1016/j. compag.2018.11.028.
techfore.2021.120880. Hunt, E.R., Daughtry, C.S.T., 2018. What good are unmanned aircraft systems for
Chen, K., Tomblin, D., 2021. Using data from Reddit, public deliberation, and surveys to agricultural remote sensing and precision agriculture? Int. J. Remote Sens. 39,
measure public opinion about autonomous vehicles. Public Opin. Q. 85, 289–322. 5345–5376. https://doi.org/10.1080/01431161.2017.1410300.
https://doi.org/10.1093/poq/nfab021. Isensee, C., Teuteberg, F., Griese, K.M., Topi, C., 2020. The relationship between
Chen, K., Duan, Z., Yang, S., 2022. Twitter as research data: tools, costs, skill sets, and organizational culture, sustainability, and digitalization in SMEs: a systematic
lessons learned. Polit. Life Sci. 41, 114–130. https://doi.org/10.1017/pls.2021.19. review. J. Clean. Prod. 275, 122944 https://doi.org/10.1016/j.
Cheng, C., Wang, L., 2021. How companies configure digital innovation attributes for jclepro.2020.122944.
business model innovation? A configurational view. Technovation 102398. https:// Jespersen, L.M., Hansen, J.P., Brunori, G., Jensen, A.L., Hols, K., Mathiesen, C.,
doi.org/10.1016/j.technovation.2021.102398. Halberg, N., Rasmussen, I.A., 2013. ICT and Social Media as Drivers of Multi- Actor
Chowdhury, A., Odame, H.H., 2013. Social media for enhancing innovation in agri-food Innovation in Agriculture – Barriers, Recommendations and Potentials. International
and rural development: current dynamics in Ontario, Canada. J. Rural Commun. Centre for Research in Organic Food Systems, Aarhus, Denmark.
Dev. 8, 97–119. Jokers, M.L., 2017. Syuzhet: Extract Sentiment and Plot Arcs from Text.
Ciarli, T., Kenney, M., Massini, S., Piscitello, L., 2021. Digital technologies, innovation, Joseph, N., Kar, A.K., Ilavarasan, P.V., Ganesh, S., 2017. Review of discussions on
and skills: emerging trajectories and challenges. Res. Policy 50, 104289. https://doi. internet of things (IoT): insights from twitter analytics. J. Glob. Inf. Manag. 25,
org/10.1016/j.respol.2021.104289. 38–51. https://doi.org/10.4018/JGIM.2017040103.
Clapp, J., Ruder, S.-L., 2020. Precision technologies for agriculture: digital farming, Juntunen, M., Ismagilova, E., Oikarinen, E.L., 2020. B2B brands on twitter: engaging
gene-edited crops, and the politics of sustainability. Glob. Environ. Polit. 20, 49–69. users with a varying combination of social media content objectives, strategies, and
https://doi.org/10.1162/glep_a_00566. tactics. Ind. Mark. Manag. 89, 630–641. https://doi.org/10.1016/j.
Cristobal-Fransi, E., Montegut-Salla, Y., Ferrer-Rosell, B., Daries, N., 2020. Rural indmarman.2019.03.001.
cooperatives in the digital age: an analysis of the internet presence and degree of Kamilaris, A., Fonts, A., Prenafeta-Boldύ, F.X., 2019. The rise of blockchain technology in
maturity of Agri-food cooperatives’ e-commerce. J. Rural. Stud. 74, 55–66. https:// agriculture and food supply chains. Trends Food Sci. Technol. 91, 640–652. https://
doi.org/10.1016/j.jrurstud.2019.11.011. doi.org/10.1016/j.tifs.2019.07.034.
Del Río Castro, G., González Fernández, M.C., Uruburu Colsa, Á., 2021. Unleashing the Kaplan, A., Haenlein, M., 2019. Digital transformation and disruption: on big data,
convergence amid digitalization and sustainability towards pursuing the sustainable blockchain, artificial intelligence, and other things. Bus. Horiz. 62, 679–681. https://
development goals (SDGs): a holistic review. J. Clean. Prod. 280, 122204 https:// doi.org/10.1016/j.bushor.2019.07.001.
doi.org/10.1016/j.jclepro.2020.122204. Karami, A., Lundy, M., Webb, F., Dwivedi, Y.K., 2020. Twitter and research: a systematic
Dey, K., Shekhawat, U., 2021. Blockchain for sustainable e-agriculture: literature review, literature review through text mining. IEEE Access 8, 67698–67717. https://doi.org/
architecture for data management, and implications. J. Clean. Prod. 316, 128254 10.1109/ACCESS.2020.2983656.
https://doi.org/10.1016/j.jclepro.2021.128254. Karami, A., Kadari, R.R., Panati, L., Nooli, S.P., Bheemreddy, H., Bozorgi, P., 2021.
Di Vaio, A., Boccia, F., Landriani, L., Palladino, R., 2020. Artificial intelligence in the Analysis of geotagging behavior: do geotagged users represent the twitter
Agri-food system: rethinking sustainable business models in the COVID-19 scenario. population? ISPRS Int. J. Geo Inf. 10 https://doi.org/10.3390/ijgi10060373.
Sustainability 12, 4851. https://doi.org/10.3390/SU12124851. Kernecker, M., Knierim, A., Wurbs, A., Kraus, T., Borges, F., 2020. Experience versus
Driessen, C., Heutinck, L.F.M., 2015. Cows desiring to be milked? Milking robots and the expectation: farmers’ perceptions of smart farming technologies for cropping
co-evolution of ethics and technology on Dutch dairy farms. Agric. Hum. Values 32, systems across Europe. Precis. Agric. 21, 34–50. https://doi.org/10.1007/s11119-
3–20. https://doi.org/10.1007/s10460-014-9515-5. 019-09651-z.
Duncan, E., Glaros, A., Ross, D.Z., Nost, E., 2021. New but for whom? Discourses of Klašnja, M., Barberá, P., Beauchamp, N., Nagler, J., Tucker, J.A., 2017. Measuring Public
innovation in precision agriculture. Agric. Hum. Values 38, 1181–1199. https://doi. Opinion with Social Media Data. Oxford University Press. https://doi.org/10.1093/
org/10.1007/s10460-021-10244-8. oxfordhb/9780190213299.013.3.
Eastwood, C., Klerkx, L., Ayre, M., Dela Rue, B., 2019. Managing socio-ethical challenges Klerkx, L., 2021. Digital and virtual spaces as sites of extension and advisory services
in the development of smart farming: from a fragmented to a comprehensive research: social media, gaming, and digitally integrated and augmented advice.
approach for responsible research and innovation. J. Agric. Environ. Ethics 32, J. Agric. Educ. Ext. 27, 277–286. https://doi.org/10.1080/
741–768. https://doi.org/10.1007/s10806-017-9704-5. 1389224X.2021.1934998.
European Commission, 2009. Preparing For Our Future: Developing a Common Strategy Klerkx, L., Begemann, S., 2020. Supporting food systems transformation: the what, why,
for Key Enabling Technologies in the EU. Communication of the European who, where and how of mission-oriented agricultural innovation systems. Agric.
Commission COM 512 final. Syst. 184, 102901 https://doi.org/10.1016/j.agsy.2020.102901.
Fielke, S., Taylor, B., Jakku, E., 2020. Digitalisation of agricultural knowledge and advice Klerkx, L., Rose, D., 2020. Dealing with the game-changing technologies of agriculture
networks: a state-of-the-art review. Agric. Syst. 180, 102763 https://doi.org/ 4.0: how do we manage diversity and responsibility in food system transition
10.1016/j.agsy.2019.102763. pathways? Glob. Food Secur. 24, 100347 https://doi.org/10.1016/j.
Firdaus, S.N., Ding, C., Sadeghian, A., 2018. Retweet: a popular information diffusion gfs.2019.100347.
mechanism – a survey paper. Online Soc. Netw. Med. 6, 26–40. https://doi.org/ Klerkx, L., Jakku, E., Labarthe, P., 2019. A review of social science on digital agriculture,
10.1016/j.osnem.2018.04.001. smart farming and agriculture 4.0: new contributions and a future research agenda.
Fountas, S., Sorensen, C.G., Tsiropoulos, Z., Cavalaris, C., Liakos, V., Gemtos, T., 2015. NJAS Wagen. J. Life Sci. 90–91, 100315 https://doi.org/10.1016/j.
Farm machinery management information system. Comput. Electron. Agric. 110, njas.2019.100315.
131–138. https://doi.org/10.1016/j.compag.2014.11.011. Krippendorff, K., 2004. Reliability in content analysis. Hum. Commun. Res. 30, 411–433.
Freidberg, S., 2017. Big food and little data: the slow harvest of corporate food supply https://doi.org/10.1111/j.1468-2958.2004.tb00738.x.
chain sustainability initiatives. Ann. Am. Assoc. Geogr. 107, 1389–1406. https://doi. Lakshmi, V., Corbett, J., 2020. How artificial intelligence improves agricultural
org/10.1080/24694452.2017.1309967. productivity and sustainability: a global thematic analysis, in. In: Proceedings of the
Galanakis, C.M., Rizou, M., Aldawoud, T.M.S., Ucak, I., Rowan, N.J., 2021. Innovations 53rd Hawaii International Conference on System Sciences, pp. 5202–5211. https://
and technology disruptions in the food sector within the COVID-19 pandemic and doi.org/10.18034/apjee.v6i2.542.
post-lockdown era. Trends Food Sci. Technol. 110, 193–200. https://doi.org/ Lanzolla, G., Lorenz, A., Miron-Spektor, E., Schilling, M., Solinas, G., Tucci, C.L., 2020.
10.1016/j.tifs.2021.02.002. Digital transformation: what is new if anything? Emerging patterns and management
Garzoni, A., De Turi, I., Secundo, G., Del Vecchio, P., 2020. Fostering digital research. Acad. Manage. Discov. 6, 341–350. https://doi.org/10.5465/
transformation of SMEs: a four levels approach. Manag. Decis. 58, 1543–1562. amd.2020.0144.
https://doi.org/10.1108/MD-07-2019-0939. Lawson, L.G., Pedersen, S.M., Sørensen, C.G., Pesonen, L., Fountas, S., Werner, A.,
Gaspar, R., Gorjão, S., Seibt, B., Lima, L., Barnett, J., Moss, A., Wills, J., 2014. Tweeting Oudshoorn, F.W., Herold, L., Chatzinikos, T., Kirketerp, I.M., Blackmore, S., 2011.
during food crises: a psychosocial analysis of threat coping expressions in Spain, A four nation survey of farm information management and advanced farming
during the 2011 European EHEC outbreak. Int. J. Human-Comput. Stud. 72, systems: a descriptive analysis of survey responses. Comput. Electron. Agric. 77,
239–254. https://doi.org/10.1016/j.ijhcs.2013.10.001. 7–20. https://doi.org/10.1016/j.compag.2011.03.002.
Gentry, J., 2015. twitteR: R Based Twitter Client. R package version 1.1.9. Leviäkangas, P., 2016. Digitalisation of Finland’s transport sector. Technol. Soc. 47,
Ghani, N.A., Hamid, S., Targio Hashem, I.A., Ahmed, E., 2019. Social media big data 1–15. https://doi.org/10.1016/j.techsoc.2016.07.001.
analytics: a survey. Comput. Hum. Behav. 101, 417–428. https://doi.org/10.1016/j. Lezoche, M., Hernandez, J.E., Díaz, Alemany, EvaPanetto, Ma Mar, Kacprzyk, J., 2020.
chb.2018.08.039. Agri-food 4.0: a survey of the supply chains and technologies for the future
Giua, C., Materia, V.C., Camanzi, L., 2020. Management information system adoption at agriculture. Comput. Ind. 117, 103187 https://doi.org/10.1016/j.
the farm level: evidence from the literature. Br. Food J. 123, 884–909. https://doi. compind.2020.103187.
org/10.1108/BFJ-05-2020-0420. Li, C., Kent, M.L., 2021. Explorations on mediated communication and beyond: toward a
Grover, P., Kar, A.K., Dwivedi, Y.K., 2020. Understanding artificial intelligence adoption theory of social media. Public Relat. Rev. 47, 102112 https://doi.org/10.1016/j.
in operations management: insights from the review of academic literature and pubrev.2021.102112.
social media discussions. Ann. Oper. Res. 1–37 https://doi.org/10.1007/s10479- Liu, B., 2015. Sentiment Analysis: Mining Opinions, Sentiments, and Emotions.
020-03683-9. Cambridge University Press.
12
Loebbecke, C., Picot, A., 2015. Reflections on societal and business model transformation Pivoto, D., Barham, B., Waquil, P.D., Foguesatto, C.R., Corte, V.F.D., Zhang, D.,
arising from digitization and big data analytics: a research agenda. J. Strateg. Inf. Talamini, E., 2019. Factors influencing the adoption of smart farming by Brazilian
Syst. 24, 149–157. https://doi.org/10.1016/j.jsis.2015.08.002. grain farmers. Int. Food Agribus. Manage. Rev. 22, 571–588. https://doi.org/
Lopez-Ridaura, S., Sanders, A., Barba-Escoto, L., Wiegel, J., Mayorga-Cortes, M., 10.22434/IFAMR2018.0086.
Gonzalez-Esquivel, C., Lopez-Ramirez, M.A., Escoto-Masis, R.M., Morales- Price, C., 2021. The online genetically modified food debate: digital food activism,
Galindo, E., García-Barcena, T.S., 2021. Immediate impact of COVID-19 pandemic science and alternative knowledges. Digital Geogr. Soc. 2, 100017 https://doi.org/
on farming systems in Central America and Mexico. Agric. Syst. 192, 103178 10.1016/j.diggeo.2021.100017.
https://doi.org/10.1016/j.agsy.2021.103178. Ramaswamy, V., Ozcan, K., 2016. Brand value co-creation in a digitalized world: an
Lu, Y., 2017. Industry 4.0: a survey on technologies, applications and open research integrative framework and research implications. Int. J. Res. Mark. 33, 93–106.
issues. J. Ind. Inf. Integr. 6, 1–10. https://doi.org/10.1016/j.jii.2017.04.005. https://doi.org/10.1016/j.ijresmar.2015.07.001.
Marshall, A., Dezuanni, M., Burgess, J., Thomas, J., Wilson, C.K., 2020. Australian Ribeiro, T.G., Barone, B., Behrens, J.H., 2016. Genetically modified foods and their social
farmers left behind in the digital economy – insights from the Australian digital representation. Food Res. Int. 84, 120–127. https://doi.org/10.1016/j.
inclusion index. J. Rural. Stud. 80, 195–210. https://doi.org/10.1016/j. foodres.2016.03.029.
jrurstud.2020.09.001. Rijswijk, K., Klerkx, L., Bacco, M., Bartolini, F., Bulten, E., Debruyne, L., Dessein, J.,
Matarazzo, M., Penco, L., Profumo, G., Quaglia, R., 2021. Digital transformation and Scotti, I., Brunori, G., 2021. Digital transformation of agriculture and rural areas: a
customer value creation in made in Italy SMEs: a dynamic capabilities perspective. socio-cyber-physical system framework to support responsibilisation. J. Rural. Stud.
J. Bus. Res. 123, 642–656. https://doi.org/10.1016/j.jbusres.2020.10.033. 85, 79–90. https://doi.org/10.1016/j.jrurstud.2021.05.003.
Mazzucato, M., Kattel, R., Ryan-Collins, J., 2020. Challenge-driven innovation policy: Rinker, T., 2017. Sentimentr: Calculate Text Polarity Sentiment.
towards a new policy toolkit. J. Ind. Compet. Trade 20, 421–437. https://doi.org/ Rose, D.C., Chilvers, J., 2018. Agriculture 4.0: broadening responsible innovation in an
10.1007/s10842-019-00329-w. era of smart farming. Front. Sustain. Food Syst. 2, 1–7. https://doi.org/10.3389/
Mellon, J., Prosser, C., 2017. Twitter and Facebook are not representative of the general fsufs.2018.00087.
population: political attitudes and demographics of British social media users. Res. Rotz, S., Duncan, E., Small, M., Botschner, J., Dara, R., Mosby, I., Reed, M., Fraser, E.D.
Polit. 4 https://doi.org/10.1177/2053168017720008, 205316801772000. G., 2019. The politics of digital agricultural technologies: a preliminary review.
Mergel, I., Bretschneider, S.I., 2013. A three-stage adoption process for social media use Sociol. Rural. 59, 203–229. https://doi.org/10.1111/soru.12233.
in government. Public Adm. Rev. 73, 390–400. https://doi.org/10.1111/ Rowan, N.J., Galanakis, C.M., 2020. Unlocking challenges and opportunities presented
puar.12021. by COVID-19 pandemic for cross-cutting disruption in agri-food and green deal
Mills, J., Reed, M., Skaalsveen, K., Ingram, J., 2019. The use of twitter for knowledge innovations: Quo Vadis? Sci. Total Environ. 748, 141362 https://doi.org/10.1016/j.
exchange on sustainable soil management. Soil Use Manag. 35, 195–203. https:// scitotenv.2020.141362.
doi.org/10.1111/sum.12485. Rutten, C.J., Velthuis, A.G.J., Steeneveld, W., Hogeveen, H., 2013. Invited review:
Miranda, J., Ponce, P., Molina, A., Wright, P., 2019. Sensing, smart and sustainable sensors to support health management on dairy farms. J. Dairy Sci. 96, 1928–1952.
technologies for Agri-food 4.0. Comput. Ind. 108, 21–36. https://doi.org/10.1016/j. https://doi.org/10.3168/jds.2012-6107.
compind.2019.02.002. Sachs, J.D., Schmidt-Traub, G., Mazzucato, M., Messner, D., Nakicenovic, N.,
Mishra, N., Singh, A., 2018. Use of twitter data for waste minimisation in beef supply Rockström, J., 2019. Six transformations to achieve the sustainable development
chain. Ann. Oper. Res. 270, 337–359. https://doi.org/10.1007/s10479-016-2303-4. goals. Nat. Sustain. 2, 805–814. https://doi.org/10.1038/s41893-019-0352-9.
Mnif, E., Mouakhar, K., Jarboui, A., 2021. Blockchain technology awareness on social Samoggia, A., Riedel, B., Ruggeri, A., 2020. Social media exploration for understanding
media: insights from twitter analytics. J. High Technol. Manage. Res. 32, 100416 food product attributes perception: the case of coffee and health with twitter data.
https://doi.org/10.1016/j.hitech.2021.100416. Br. Food J. 122, 3815–3835. https://doi.org/10.1108/BFJ-03-2019-0172.
Moe, W.W., Schweidel, D.A., 2017. Opportunities for innovation in social media Schaefers, T., Falk, T., Kumar, A., Schamari, J., 2021. More of the same? Effects of
analytics. J. Prod. Innov. Manag. 34, 697–702. https://doi.org/10.1111/ volume and variety of social media brand engagement behavior. J. Bus. Res. 135,
jpim.12405. 282–294. https://doi.org/10.1016/j.jbusres.2021.06.033.
Mondal, P., Basu, M., 2009. Adoption of precision agriculture technologies in India and Shawky, S., Kubacki, K., Dietrich, T., Weaven, S., 2020. A dynamic framework for
in some developing countries: scope, present status and strategies. Prog. Nat. Sci. 19, managing customer engagement on social media. J. Bus. Res. 121, 567–577. https://
659–666. https://doi.org/10.1016/j.pnsc.2008.07.020. doi.org/10.1016/j.jbusres.2020.03.030.
Mondejar, M.E., Avtar, R., Diaz, H.L.B., Dubey, R.K., Esteban, J., Gómez-Morales, A., Shepherd, M., Turner, J.A., Small, B., Wheeler, D., 2020. Priorities for science to
Hallam, B., Mbungu, N.T., Okolo, C.C., Prasad, K.A., She, Q., Garcia-Segura, S., overcome hurdles thwarting the full promise of the ‘digital agriculture’ revolution.
2021. Digitalization to achieve sustainable development goals: steps towards a smart J. Sci. Food Agric. 100, 5083–5092. https://doi.org/10.1002/jsfa.9346.
green planet. Sci. Total Environ. 794, 148539 https://doi.org/10.1016/j. Silvestre, B.S., Ţîrcă, D.M., 2019. Innovations for sustainable development: moving
scitotenv.2021.148539. toward a sustainable future. J. Clean. Prod. 208, 325–332. https://doi.org/10.1016/
Monteiro, J., Barata, J., 2021. Artificial intelligence in extended Agri-food supply chain: j.jclepro.2018.09.244.
a short review based on bibliometric analysis. Proc. Comput. Sci. 192, 3020–3029. Sinha, N., Singh, Pragati, Gupta, M., Singh, Pratibha, 2020. Robotics at workplace: an
https://doi.org/10.1016/j.procs.2021.09.074. integrated twitter analytics – SEM based approach for behavioral intention to accept.
Montgomery, K., Grier, S., Chester, J., Dorfman, L., 2011. Food Marketing in the Digital Int. J. Inf. Manag. 55, 102210 https://doi.org/10.1016/j.ijinfomgt.2020.102210.
Age: A Conceptual Framework and Agenda for Research. Statista, 2021. Leading Countries Based on Number of Twitter Users as of October 2021
Mostafa, M.M., 2013. More than words: social networks’ text mining for consumer brand (in Millions) [WWW Document]. URL. https://www.statista.com/statistics/242606/
sentiments. Expert Syst. Appl. 40, 4241–4251. https://doi.org/10.1016/j. number-of-active-twitter-users-in-selected-countries/ (accessed 10.23.21).
eswa.2013.01.019. Stevens, T.M., Aarts, N., Termeer, C.J.A.M., Dewulf, A., 2016. Social media as a new
Moysiadis, V., Sarigiannidis, P., Vitsas, V., Khelifi, A., 2021. Smart farming in Europe. playing field for the governance of agro-food sustainability. Curr. Opin. Environ.
Comput. Sci. Rev. 39, 100345 https://doi.org/10.1016/J.COSREV.2020.100345. Sustain. 18, 99–106. https://doi.org/10.1016/j.cosust.2015.11.010.
Norton, G.W., Alwang, J., 2020. Changes in agricultural extension and implications for Storm, H., Baylis, K., Heckelei, T., 2020. Machine learning in agricultural and applied
farmer adoption of new practices. Appl. Econ. Perspect. Pol. 42, 8–20. https://doi. economics. Eur. Rev. Agric. Econ. 47, 849–892. https://doi.org/10.1093/erae/
org/10.1002/aepp.13008. jbz033.
Ofori, M., El-Gayar, O., 2021. Drivers and challenges of precision agriculture: a social Sullivan, C., Wong, I., Adams, E., Fahim, M., Fraser, J., Ranatunga, G., Busato, M.,
media perspective. Precis. Agric. 22, 1019–1044. https://doi.org/10.1007/s11119- McNeil, K., 2021. Moving faster than the COVID-19 pandemic: the rapid, digital
020-09760-0. transformation of a public health system. Appl. Clin. Inform. 12, 229–236. https://
Paniagua, J., Korzynski, P., Mas-Tur, A., 2017. Crossing borders with social media: online doi.org/10.1055/s-0041-1725186.
social networks and FDI. Eur. Manag. J. 35, 314–326. https://doi.org/10.1016/j. Swani, K., Brown, B.P., Milne, G.R., 2014. Should tweets differ for B2B and B2C? An
emj.2016.09.002. analysis of fortune 500 companies’ twitter communications. Ind. Mark. Manag. 43,
Parra-López, C., Reina-Usuga, L., Carmona-Torres, C., Sayadi, S., Klerkx, L., 2021. Digital 873–881. https://doi.org/10.1016/j.indmarman.2014.04.012.
transformation of the agrifood system: quantifying the conditioning factors to inform Tey, Y.S., Brindal, M., 2012. Factors influencing the adoption of precision agricultural
policy planning in the olive sector. Land Use Policy 108, 105537. https://doi.org/ technologies: a review for policy implications. Precis. Agric. 13, 713–730. https://
10.1016/j.landusepol.2021.105537. doi.org/10.1007/s11119-012-9273-6.
Petersen, K., Gerken, J.M., 2021. #Covid-19: an exploratory investigation of hashtag Torky, M., Hassanein, A.E., 2020. Integrating blockchain and the internet of things in
usage on twitter. Health Policy 125, 541–547. https://doi.org/10.1016/j. precision agriculture: analysis, opportunities, and challenges. Comput. Electron.
healthpol.2021.01.001. Agric. 178, 105476 https://doi.org/10.1016/j.compag.2020.105476.
Pfeiffer, J., Gabriel, A., Gandorfer, M., 2021. Understanding the public attitudinal Tortora, D., Chierici, R., Farina Briamonte, M., Tiscini, R., 2021. ‘I digitize so I exist’.
acceptance of digital farming technologies: a nationwide survey in Germany. Agric. Searching for critical capabilities affecting firms’ digital innovation. J. Bus. Res. 129,
Hum. Values 38, 107–128. https://doi.org/10.1007/s10460-020-10145-2. 193–204. https://doi.org/10.1016/j.jbusres.2021.02.048.
Phillips, T., Klerkx, L., Mcentee, M., 2018. An investigation of social Media’s roles in Trivelli, L., Apicella, A., Chiarello, F., Rana, R., Fantoni, G., Tarabella, A., 2019. From
knowledge exchange by farmers. In: 13th European IFSA Symposium. precision agriculture to industry 4.0: unveiling technological connections in the
Phillips, T., McEntee, M., Klerkx, L., 2021. An investigation into the use of social media agrifood sector. Br. Food J. 121, 1730–1743. https://doi.org/10.1108/BFJ-11-2018-
for knowledge exchange by farmers and advisors. Rural Extens. Innov. Syst. J. 17, 0747.
1–13. Vaccari, C., Valeriani, A., Barberá, P., Bonneau, R., Jost, J.T., Nagler, J., Tucker, J., 2013.
Pindado, E., Barrena, R., 2020. Using twitter to explore consumers’ sentiments and their Social media and political communication. A survey of twitter users during the 2013
social representations towards new food trends. Br. Food J. 123, 1060–1082. Italian general election. Riv. Ital. Sci. Polit. 43, 381–410. https://doi.org/10.1426/
https://doi.org/10.1108/BFJ-03-2020-0192. 75245.
13
Veltri, G.A., Atanasova, D., 2015. Climate change on twitter: content, media ecology and World Bank, 2021. World Development Report 2021: Data for Better Lives.
information sharing behaviour. Public Underst. Sci. 26, 721–737. https://doi.org/ World Economic Forum, 2020. The Future of Jobs Report. Geneva, Switzerland.
10.1177/0963662515613702. Wyckoff, A., Pilat, D., 2017. Key Issues for Digital Transformation in the G20.
Verhoef, P.C., Broekhuizen, T., Bart, Y., Bhattacharya, A., Qi Dong, J., Fabian, N., Yadav, V.S., Singh, A.R., Raut, R.D., Mangla, S.K., Luthra, S., Kumar, A., 2022. Exploring
Haenlein, M., 2021. Digital transformation: a multidisciplinary reflection and the application of industry 4.0 technologies in the agricultural food supply chain: a
research agenda. J. Bus. Res. 122, 889–901. https://doi.org/10.1016/j. systematic literature review. Comput. Ind. Eng. 169, 108304 https://doi.org/
jbusres.2019.09.022. 10.1016/j.cie.2022.108304.
Vidal, L., Ares, G., Machín, L., Jaeger, S.R., 2015. Using Twitter data for food-related Yang, M., Fu, M., Zhang, Z., 2021a. The adoption of digital technologies in supply chains:
consumer research: a case study on “what people say when tweeting about different drivers, process and impact. Technol. Forecast. Soc. Chang. 169, 120795 https://doi.
eating situations”. Food Qual. Prefer. 45, 58–69. https://doi.org/10.1016/j. org/10.1016/j.techfore.2021.120795.
foodqual.2015.05.006. Yang, K.-C., Pierri, F., Hui, P.-M., Axelrod, D., Torres-Lugo, C., Bryden, J., Menczer, F.,
Vinuesa, R., Azizpour, H., Leite, I., Balaam, M., Dignum, V., Domisch, S., Felländer, A., 2021b. The COVID-19 infodemic: twitter versus facebook. Big Data Soc. 8 https://
Langhans, S.D., Tegmark, M., Fuso Nerini, F., 2020. The role of artificial intelligence doi.org/10.1177/20539517211013861, 205395172110138.
in achieving the sustainable development goals. Nat. Commun. 11, 1–10. https:// Yoon, C., Lim, D., Park, C., 2020. Factors affecting adoption of smart farms: the case of
doi.org/10.1038/s41467-019-14108-y. Korea. Comput. Hum. Behav. 108, 106309 https://doi.org/10.1016/j.
Warner, K.S.R., Wäger, M., 2019. Building dynamic capabilities for digital chb.2020.106309.
transformation: an ongoing process of strategic renewal. Long Range Plan. 52, Zaglia, M.E., 2013. Brand communities embedded in social networks. J. Bus. Res. 66,
326–349. https://doi.org/10.1016/j.lrp.2018.12.001. 216–223. https://doi.org/10.1016/j.jbusres.2012.07.015.
Wathes, C.M., Kristensen, H.H., Aerts, J.M., Berckmans, D., 2008. Is precision livestock Zambon, I., Cecchini, M., Egidi, G., Saporito, M.G., Colantoni, A., 2019. Revolution 4.0:
farming an engineer’s daydream or nightmare, an animal’s friend or foe, and a industry vs. agriculture in a future development for SMEs. Processes 7, 36. https://
farmer’s panacea or pitfall? Comput. Electron. Agric. 64, 2–10. https://doi.org/ doi.org/10.3390/pr7010036.
10.1016/j.compag.2008.05.005. Zeng, Z., Li, S., Lian, J.W., Li, J., Chen, T., Li, Y., 2021. Switching behavior in the
Weiss, S.M., Indurkhya, N., Zhang, T., Damerau, F.J., 2005. Text Mining: Predictive adoption of a land information system in China: a perspective of the
Methods for Analyzing Unstructured Information. Springer, New York. https://doi. push–pull–mooring framework. Land Use Policy 109, 105629. https://doi.org/
org/10.1007/978-0-387-34555-0. 10.1016/j.landusepol.2021.105629.
Wilson, T., Wiebe, J., Hoffmann, P., 2009. Recognizing contextual polarity: an Żmija, K., Fortes, A., Tia, M.N., Šūmane, S., Ayambila, S.N., Żmija, D., Satoła, Ł.,
exploration of features for phrase-level sentiment analysis. Comput. Linguist. 35, Sutherland, L.A., 2020. Small farming and generational renewal in the context of
399–433. https://doi.org/10.1162/coli.08-012-R1-06-90. food security challenges. Glob. Food Secur. 26, 100412 https://doi.org/10.1016/j.
Wolfert, S., Ge, L., Verdouw, C., Bogaardt, M.J., 2017. Big data in smart farming – a gfs.2020.100412.
review. Agric. Syst. 153, 69–80. https://doi.org/10.1016/j.agsy.2017.01.023.
14

1 s2.0 S0308521X22001561 Main

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 s2.0 S0308521X22001561 Main

Uploaded by

Copyright:

Available Formats

Agricultural Systems 203 (2022) 103520

Contents lists available at ScienceDirect

New trends in the global digital transformation process of the agri-food

• The adoption of new digital technolo

Fig. 1. Proposed Twitter analytics framework.

Fig. 2. Tweets posted during the analysed period.

It was followed by India, the UK and Nigeria, with means of 14.3%,

We further performed word analysis in the four most active coun

2-word comb. Freq % 2-word comb. Freq % 2-word comb. Freq %

Cloud computing AI Machine learning

2-word comb. Freq % 2-word comb. Freq % 2-word comb. Freq %

Advanced farming solutions to improve productivity, efficiency and 2.03

5.2. Limitations and future research org/10.1016/j.agsy.2022.103520.

You might also like

1 s2.0 S0308521X22001561 Main

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 s2.0 S0308521X22001561 Main

Uploaded by

Copyright:

Available Formats

Agricultural Systems 203 (2022) 103520

Contents lists available at ScienceDirect

New trends in the global digital transformation process of the agri-food

• The adoption of new digital technolo­

Fig. 1. Proposed Twitter analytics framework.

Fig. 2. Tweets posted during the analysed period.

It was followed by India, the UK and Nigeria, with means of 14.3%,

We further performed word analysis in the four most active coun­

2-word comb. Freq % 2-word comb. Freq % 2-word comb. Freq %

Cloud computing AI Machine learning

2-word comb. Freq % 2-word comb. Freq % 2-word comb. Freq %

Advanced farming solutions to improve productivity, efficiency and 2.03

5.2. Limitations and future research org/10.1016/j.agsy.2022.103520.

You might also like

• The adoption of new digital technolo

We further performed word analysis in the four most active coun