Need and Relevance of Common
Vocabularies and Ontologies in IoT
Domain
Arunima Sharma and Ramesh Babu Battula
Abstract Data plays a foremost part in Future Communication. Internet of Things
encompasses various sources of data of different type, by different name. The
complexity and variance in data increase the criticalness of the model. Same data
has different meaning in different applications. These variances make it difficult to
apply any query on data. To resolve the issues in naming of data values and to create
consistency Vocabulary and Ontologies are must. Data gathered on a big platform
e.g. for IoT via cloud need to manage data in proper manner, which is spread all
over the internet on different servers. The features and related research is explained
in this chapter, having a brief introduction with application areas.
Keywords Semantic web · Ontology · Vocabulary · Internet of thing · IoT · Smart
city
1 Introduction
Internet of Things (IoT) brings a heterogeneus of technologies and devices together
to simplify the basic life of human beings. Smart Home, Traffic, Healthcare etc. are
some of the major areas using different IoT components. On basis of applications
we can consider IoT as a complex group of several devices, sensors and people who
interacts with each other and controlled by a centre control station. This interaction
or communication generates a large amount of data gathered from several source in
different formats. This data has various formats of data related to each other. Due
to complex relationship and variation this data is considered as heterogeneous data.
Heterogeneous data is complex with perspective of management, access and collection. On web, data is linked together on basis of some relationship and similarity. The
extension of World Wide Web which deal with related web pages, and extensions of
similar data is known as Semantic Web.
A. Sharma (B) · R. B. Battula
Department of Computer Science and Engineering, Malviya National Institute of Technology,
Jaipur, India
e-mail: aru.92@rediffmail.com
© Springer Nature Switzerland AG 2021
R. Pandey et al. (eds.), Semantic IoT: Theory and Applications, Studies in Computational
Intelligence 941, https://doi.org/10.1007/978-3-030-64619-6_6
133
134
A. Sharma and R. B. Battula
Complex linking in words on internet in interlinked with each other and creates a
web of connections between words related to each other. Vocabulary and Ontology
are two main characteristics of words which play most important role. The IoT
assumes an ever-expanding job in empowering brilliant city applications. A philosophy based semantic methodology can help improve interoperability between an
assortment of IoT-created just as reciprocal information expected to drive these
applications. While various metaphysics indexes exist, utilizing them for IoT and
brilliant city applications require critical measure of work. Semantic advancements
propose an appropriate methodology for interoperability by sharing normal vocabularies, and furthermore empowering interoperable portrayal of construed information. IoT testbed suppliers have as of late began to add semantics to their structures
permitting the creation of the semantic Sensor Web, which is an expansion of the
current Web in which data is given all around characterized which means, empowering machine-to-machine interchanges and cooperations between items, gadgets
and individuals.
Semantics generally model the space ideas in incredible detail. Despite the fact that
they can be applied for questioning nearly anything about articles, these unpredictable
models are regularly hard to execute and utilize, particularly by non-specialists.
They request extensive handling assets and along these lines they are viewed as
unacceptable for compelled situations. Rather, IoT models ought to think about the
limitations and dynamicity of the IoT situations, particularly with the new pattern
towards incorporating semantics in obliged gadgets for example, M2M doors or cell
phones. Simultaneously they have to show the connections and ideas that speak to
and permit interoperability between IoT elements.
In this manner, expressiveness versus intricacy is a test. It is essential to take
note of that semantic models are most certainly not final results. They are ordinarily just piece of an answer also, ought to be straightforward to the end client. The
semantic comment models ought to be offered with compelling strategies, Programming interface’s and instruments to process the semantics so as to extricate significant data from crude information. Question techniques, AI, thinking and information
investigation methods ought to have the option to successfully utilize these semantics. Semantic demonstrating is just the underlying piece of the entire plan, and it
needs to consider how the models will be utilized; how the explained information
will be recorded and questioned with continuous information; and how to make the
distribution reasonable for obliged conditions and enormous scope organizations
when applications frequently require low inactivity and handling time. To permit a
typical jargon to interoperate between various frameworks a scientific classification
is required to portray the estimations of the gadgets as far as the amount sorts and
units, for example, temperature and degrees Celsius.
Need and Relevance of Common Vocabularies and Ontologies …
135
2 Vocabulary
Vocabulary is the knowledge regarding word. It is the basic information required
about a word.
Controlled vocabularies provide a way to arrange data for successive information
retrieval. Indexing, Heading, Thesauri, and other are collection of controlled data.
This data is well defined selected for a category to belong on basis of knowledge,
natural language and some restrictions. It is well formed set of information so that
any query on data will give corrects outcome. It uses tagging system for words to
get connected and retrieval becomes easy.
2.1 Vocabulary Tools
1. Subject Heading
Subject heading helps to create the catalogue of data in categories and subcategories easily. It considers the whole document in pre-coordinated order.
2. Thesauri
It is collection of words well organized in a sequence. It covers a specific portion
of document in direct order.
3 Ontology
Ontology is derived from Greek words means being logical discourse, in general like
a dictionary. Ontology consist all the relations, rules, similarity and dissimilarity. It is
the formal representation, categorization, and domain of data. All the ontologies are
used to resolve the complex relationships. These ontologies are limited in number
till there is no new requirement raised for a different sub category.
Ontologies, for example, OWL-S and low-level determinations, for example, the
TD or the oneM2M Base Ontology can be utilized together to depict IoT/WoT frameworks, cultivating interoperability. OWL-S helps programming operators to find the
web administration that will satisfy a particular need. When found, OWL-S gives
the vital language builds to depict how to conjure the administration. It permits
portraying inputs what’s more, yields. Because of the elevated level portrayal of
the administration, it is conceivable to form numerous administrations to perform
increasingly complex errands. In OWL-S, administration depiction is composed into
four regions: the procedure model, the profile, the establishing and the administration. In particular, the procedure model depicts how a help plays out its undertakings.
It incorporates data about sources of info, yields, preconditions and results.
136
A. Sharma and R. B. Battula
Additionally, the oneM2M Global Initiative characterizes a norm for machine-tomachine correspondence interoperability at the semantic level, the one M2M Base
Ontology, which is a significant level metaphysics intended to encourage interoperability among different ontologies utilizing equivalences and arrangements. The TD
is a focal structure obstruct in the W3C Web of Things (WoT) and can be considered as
the passage purpose of a Thing. The TD comprises of semantic metadata for the Thing
itself, an association model dependent on WoT’s Properties, Actions, and Events
worldview, a semantic pattern to make information models machine-justifiable, and
highlights for Web Linking to communicate relations among Things”.
Both oneM2M Base Ontology and the metaphysics characterized in TD take
a stab at interoperability among different IoT applications and stages, every one
covering a huge arrangement of utilization cases, so there is additionally a work in
progress to adjust the oneM2M philosophy to the TD cosmology. Symmetrical to
these ontologies, the SOSA/SSN philosophy is a cosmology for depicting sensors
and their perceptions. Among other ideas, it characterizes the class SOSA: Procedure, which “is a reusable work process, convention, plan, calculation, or then again
computational strategy that can be utilized, among others, to indicate how a perception movement has been made, can be either a record of how the activation has been
performed or then again a portrayal of how to associate with an actuator (i.e., the
formula for performing incitations).
Things utilizing ontologies in a way that permits us to legitimately convey the
conduct usage. Work planned for displaying the conduct of Things utilizing FSMs and
Web Ontology Language (OWL) exists in the writing. Unified Modeling Language
(UML) FSMs utilizing OWL, playing out a very nearly balanced interpretation
between UML ideas and OWL classes. In spite of the fact that their planning from
UML to OWL takes into account a more machine-lucid data structure, its intricacy
makes it unconventional to use. UML is utilized to indicate stage free route guides
in web applications. They use OWL to portray a model for FSMs which fills in as a
meta-model for semantic web portrayals of the route guides on the Semantic Web.
There likewise exists some logical writing gave not exclusively to make a model to
communicate the conduct of a help yet in addition to decipher furthermore, execute
the conduct. The point is to build up a FSM that a unique server can peruse and mean
executable elements. These executable substances are executed later by robots. They
effectively fulfill their goal, their OWL FSM is space explicit and it incorporates
properties that illuminate the multifaceted nature of their utilization case yet make
the FSM excessively complex.
3.1 Components of Ontology
1. Individuals
Initial object considered in starting phase of classification are known as individuals. It is the basic components of ontology which comprise real objects such
Need and Relevance of Common Vocabularies and Ontologies …
137
Vehicle
Two
Wheeler
Bicycle
Scooter
Four
Wheeler
Car
Six
Wheeler
Jeep
Truck
Bus
Fig. 1 Example of ontology
as individuals, amphibians, books, plants, along with nonconcrete objects such as
numbers and words. The overall purpose of ontology is to offer a way to categorizing
entities, in different classes grouped together or correlated with each other. In Fig. 1
car, bicycle, truck are individuals.
2. Classes
Categories in which Individuals are classified are known as classes. Classes are
considered as type, category, sort, kind and extension. Classes are considered as
abstract objects that are defined by aspect values or constraints to be member of a
class. In Fig. 1 Vehicle is a class having two categories based on number of wheels.
Classes have to fulfil criteria to belong to a specific group. Class can be a collection
of other classes.
Ontologies differ on the basis of conditions related to some conditions like classes
can encompass supplementary classes, or can be applicable to itself, and if there is
a universal class, etc. Occasionally some restrictions are also applied on classes
to avoid inconsistencies. A class can be extensional if it is categorized only by its
association or class is identical else it is an intentional class. Extensional classes have
a systematic format and try to avoid any inconsistency in data.
Most of the upper ontologies are considered as intentional classes having only
necessary conditions for member association in a class. The amalgamation of
compulsory and necessary conditions within a class is considered as fully defined.
3. Attributes
Features or Properties of object which help them to categorize in particular type
of class are known as attributes. Individual objects in ontology can be related to other
things, on basis of some features or portions. These connected properties are their
attributes, even though there can be present independent objects also. Every single
attribute can be reflected as a class or individual or a sub class. The relationship among
individuals can be explained on basis of attributes easily with their similarity and
differences. For example numbers of wheels are classifying the vehicles in different
categories as shown in Fig. 1. An attribute can be present as a complex data form, i.e.
138
A. Sharma and R. B. Battula
on basis of wheels, sources, colour many vehicles get classify in different categories
but they are having some common features and inter relationships.
Ontologies are only strong if attributes are related to other attributes. If that is not
the situation, then either taxonomy or controlled vocabulary is used.
4. Relations
The way in which objects or classes are connected with each other based on a
property is known as Relation between objects or classes. Relationships (relations)
among entities in ontology identify how entities are interrelated to each other. For
example, in Fig. 1 both truck and bus are six wheeler vehicles. Set of associations
defines the semantics of the area. An important category of relation is the subsumption relation. This relationship describes which object belongs to which class.
This helps to create a tree structure of objects which represents relation among objects
having common attributes and properties. Another one is mereology relation, which
represents association in a class or classes of objects. It uses a directed acyclic
graph for data representation. Relation can be categories for relations among classes,
individuals, single individual and one class, a object and a collection of objects and
collections
5. Function terms
Function terms are structure which is get developed on basis of a relation among
individuals. It
6. Restrictions
A limiting condition which used to satisfied for development of range and domain
for individuals is Restriction on Individual. It restricts objects in a certain type of
concept.
7. Rules
Set of some conditions need to be fulfil by individual to qualify the minimum
requirement to create a relation or to belong to a class. It is a logical statement
8. Axioms
Prior Knowledge about the individuals, help to create Axioms. Axioms are set
of rules known before and based on learned data. These assumptions plays most
important role. The prior knowledge is necessary about the application to decide the
axioms in ontology. These will help to initialize basic known category for data. This
will reduce the data processing time and complexity of the processing.
9. Events
The change of state of one individual to another based on some logic is known as
event. It shows the transaction among the relationship or attributes of individuals on
a input.
Need and Relevance of Common Vocabularies and Ontologies …
139
3.2 Types of Ontology
Ontologies help to represent information and theories in different forms at different
levels of abstraction. These different domain ontologies are every so often incompatible with other ontologies. Different types of ontologies are as follow
1. Domain specific Ontology
Individuals belong to a particular type of class comes under this category. The
Domain specific ontology is classified on such sets which have no intersection and
have different meanings for different classes. For example, “name”, can be “full
name”, “last name”, “first name”… all these have different meanings when they get
connected to different classes. Merging of different ontologies is complex and has
to be done on basis of a common platform.
2. Hybrid Ontology
Hybrid is mix up of different ontologies in one which doesn’t come in remaining
categories. It is the most complex type of data which need variations in Axioms with
changes of class for an individual.
3. Upper Ontology
The general relationships among individuals which are well known and accepted
belong to upper ontology. It considers linguistic rules and learned domain for
categorization of individuals.
Ontologies need to be designed for significant applications according to user
requirements. It is required to be flexible enough for modification and reuse. Linked
data, taxonomies, interoperability, axioms, vocabulary, need to be predefined and
easy to use. Upper ontologies epitomize mutual notions and relations in range of
domains and Domain ontologies characterize conceptions and relations of a specific.
3.3 IoT Ontologies
A cosmology is characterized as “a formal, unequivocal detail of a common conceptualization” and is utilized to speak to information inside a space as a lot of ideas
identified with one another. There are four fundamental parts that form a cosmology:
Classes, relations, qualities and people. Classes are the primary ideas to depict.
Each class can have one or a few kids, known as subclasses, used to characterize
increasingly explicit ideas. Classes and subclasses have qualities that speak to their
properties and attributes. People are occasions of classes or their properties. At long
last, relations are the edges that interface all the introduced segments.
IoT-based genuine world to be isolated into 3 layers: a physical layer, i.e., things;
an data layer, i.e., information and metadata about information given by things; and a
useful layer containing administrations gave by things. To coordinate our vision of the
140
A. Sharma and R. B. Battula
genuine world and its portrayal by the Internet of Things, fabricating a metaphysics
that really models every one of the three layers. Indeed, the physical layer is spoken
to by a Device Metaphysics. The data and administration layers are spoken to by a
(Physics and Mathematics Domain Ontology and Estimation Models Ontology. To
portray the ontologies all the more decisively:
1. Gadget Ontology
The Device Ontology models real equipment gadgets that may exist in the system.
It tends to be viewed as the gadget depiction storehouse that can be gotten to for
disclosure.
2. Space Ontology
The (Physics and Mathematics) Space Ontology models data about genuine world
physical ideas and their relations among one another. It tends to be viewed as the
fundamental archive to access for administration creation.
3. Estimation Ontology
The Estimation Ontology contains data about various estimation models (“straight
introduction”, “Kalman channel”, “gullible Bayesian learning”, and so on.), the
conditions that drive them, the administrations that execute them, etc. It tends to
be for the most part viewed as the archive portraying the gadget’s nature of administration, and gives data expected to support structure. This cosmology to be utilized
as a source of perspective by any middleware or application requiring IoT administrations, i.e., administrations gave by genuine world things. Those administrations,
much of the time, produce inexact be that as it may, never 100% exact results.
Most existing philosophy work concentrated on demonstrating either gadgets as
done, e.g., in MMI cosmology 2, or material science independently. The curiosity of
our methodology is that it joins and takes favorable circumstances of the three ontologies by connecting, all together, the area of information for detecting, activating, and
handling assignments and this present reality portrayal through IoT administrations,
that know about their condition. A significant commitment is the degree of deliberation at which things, permit clients to portray gadgets in an expressive way while
as yet staying away from complex subtleties. Truth be told, as target adaptability,
consider straightforwardness in displaying information to be a basic models.
3.4 Advantages of Ontology
One of the main advantages of ontologies is that, by having information only
regarding the necessary relationships between objects helps to identify information
about data easily. Such relationship makes it is easy to implement for graphical data.
Ontologies function helps to depict the human perspective about data and interlinked
concepts among objects. In addition ontologies make available a extra rational and
tranquil navigation for operators in the ontology organization.
Need and Relevance of Common Vocabularies and Ontologies …
141
Ontologies are tranquil to outspread as associations and notion equivalently easy
to connect with the present ontologies. It also provides the meaningful significance
of datum in different format; even data is unstructured, semi-structured or structured, making data amalgamation simpler with text quarrying, and information based
analysis.
3.5 Restrictions of Ontology
Ontologies offer an opulent group of tools and techniques for information modelling,
but it comes by way of some limitations also. One such restriction is the existing property paradigms. For instance, while providing dominant class, the most recent version
of the Web Ontology Language—OWL2 has limited set of property paradigms.
Another limitation is that they specify how data has to be structured and avoid
addition of extra data. Frequently, data introduced from a new source would be
operationally inconsistent with the constraints set. Therefore, this new data need to
be modified before being incorporated with existing data.
4 Need of Vocabulary and Ontology
Efficient data storage is must needed to keep relevant multimedia data connected and
organized. This data is having lots of variance with large volume and velocity. IoT
is going to gather data from different applications developed and used by different
people is going to have inconsistency in naming methodology. If a similar vocabulary
is used it become easy to gather relevant data. Query processing and decision making
processes will also become an easy task for end users to gather important information
and used it further.
5 Ontology Quality Methodology
For evaluation and validation ontologies, ontology quality methodology [1] is used.
It has following categories to measure the quality of data
(1) Serialization
Serialization is the sequential process to convert data of an individual object into
form of binary values. Bit data is easy to save and send during communication. Its
main aim is to save all different states of data, which make its serial retrieval easy. The
inverse process of serialization is called deserialization. Serialization allows user to
save sequential states of an individual object and re-construct it also by retracing. It
provide data storage and data exchange simple among distributed applications. With
142
A. Sharma and R. B. Battula
flexibility of serialization, user can perform different actions like sending the object
by using a web service, object passing from one domain to another or through a firewall etc. Serialization helps to maintain security or user-specific shared information
across different heterogeneous applications.
(2) Syntactic validation
Syntax validation is the method of to check the syntax of a program is free of
any error. There are so many tools available to check the syntax of language. These
validators work both online and offline. There are syntax validators, know as linters,
for programming language
Syntax validators can check syntax, fading common errors such as dividing by
zero. Syntax validators highlight code style. Most of the search engines on web use
online code checkers.
(3) Interlinking
Linked data is structured data which is inter-related with each other so it becomes
more valuable through semantic queries. It is based on standard Web technologies
but rather than using those to use as web pages only it used to share information
which is automatically accessed by digital devices. One part of the visualization of
linked data is Internet which is source of global database.
(4) Documentation
Data documentation helps to keeping data organised easy to collect. The resulting
data libraries include information and method how to process it. Brief and clear documentation help users to understand the framework of data. It increases the probability
of data usage.
(5) Visualization
Ontology Visualization is representation of data in form of graph as node link
diagram having edges and vertices for visual representation of data. Usually vertices
contain semantic information and attributes. Sometimes external features are also
used, like colour shape and codes to support user visualization flexibility.
(6) Availability of resources
The resource availability describe how ontologies availability changes over time
and how user and developer observe those changes in ontologies. These attributes
include longevity and observability. Attributes of individuals captures whether the
availability of ontology is directly observable, partially observable, or unobservable
to various entities.
(7) Discoverability
Ontologies can be used to determine information from Big Data. Based on these
distinct projections, efforts can be on the way to exploration of the likelihood to
use ontological knowledge for the requirement of Big data and meta data as a
underpinning to proficiently determine useful statistics for exploration.
Need and Relevance of Common Vocabularies and Ontologies …
143
(8) Ontology Design
What’s, why’s and where’s organise the ontology in data set to identify the best
place for data. Ontology designers need to find correct place to classify data in
ontology with its relationship with rest. Matching problems doesn’t provide optimal
solution which makes need of ontology design. Domain and task are two major
entities in designing of ontologies. Domain establishes a relationship and information
of individuals and task makes them identify able in the set.
Selection of Ontology has to be based on some parameters are as follow:
1. Goal of ontology
Sharing common understanding of the structure of data is one of the major goals
of ontology. For example, if different Web sites contain medical data it becomes
difficult to share it or correlate it with other sites data. If these Web domains share
and distribute the similar ontology, data extraction and aggregation becomes easy.
The users can practice this combined statistics to reply enquiries or as feedback into
other applications.
Permitting reprocess of domain information was unique and powerful services
after recent outpouring in ontology exploration. For instance, models used for representation of dissimilar time in different domains. This comprises the concepts of time
pauses among domains, points in time, comparative time measurement, and much
more. If we requires a large ontology, integration of several present ontologies help
to create better and enormous province.
Production of unambiguous domain norms creates it easy on the way to alter if
there is any changes occurred. Separating the domain describe mechanisms rendering
to the mandatory measurement and implement it independently.
2. Size
Large size of ontology increases its efficiency as well as complexity. The small
size ontologies are not enough to perform any operation, so it gets integrated with
parent class but still remain as an individual class. The big ontology requires optimal
search and update algorithms whereas classification in such data sets is more accurate.
Interrelationship makes it difficult to manage and find a relation individually among
objects or classes.
3. Documentation
Proper documentation of ontology makes it easy to use and access. Usually for
developer it is easy to extend and integrate different data values with the help of
documentation. The proper management of ontology make easy to identify an error.
4. Avability on Web
Avability of ontology on web make it easy to update frequently having more
precise and updated knowledge. The linking of data also becomes easy to categorized data because of it. If an unknown user having less or no information will edit
data it will decrease its performance and quality. Identification of such faults is too
144
A. Sharma and R. B. Battula
difficult in such cases. Avability on web make ontologies to easily available anytime
anywhere. The queries and correlation of data become easy in this case. It will remain
updated in real time with multi user accessibility.
5. Popularity
Popularity is the most important aspect to improve the quality of words and it plays
an important role in data set. Unpopular words get vanished from daily conversation
and rarely used. The elimination of this data is not a good decision because it helps
to identify miscellaneous hidden information. But it has to give low priority for
optimal identification. Most of the least used words represent unauthorized data with
respective of forensics. Arrangement of ontologies on basis of popularity helps to
make frequently used data easily available. It makes operations less time consuming.
System optimality gets increased with this feature.
6. Maintenance with time
Maintenance of data allows speeding up query process by illuminating the inherent
relations which are authenticated and approved through the semantics of ontology.
The complication of reasoning within ontology, is in this manner moved from demand
time to apprise time.
7. Meta data Vocabulary
“Data about Data” or set of words denoting the features of resources is known
as Meta data. Ontology is a metadata vocabulary. Generally, metadata vocabularies
are domain precise. Controlled vocabularies are homogeneous and ordered words
provide a dependable way to define data. Metadata designers allocate terms from
vocabularies to simplify information recovery.
Commonly vocabularies contain subject headings, lists, files, and thesauri.
Controlled vocabularies can be organised in alphabetical lists with a hierarchical
arrangement of data. Thesauri also comprise synonyms, associated data, scope and
notes, data antiquity, alternative idioms, or numerical data. Ontologies include even
more specification, such as description, concepts, hierarchy and relationships with
other values.
6 Application
Smart City [2] requires interoperable system, to process data securely, and manages
services. Secure data access will reduce the cost of operation. Some famous Ontologies for Smart Cities [1] are KM4City (Knowledge Model for City), Semantic Traffic
Analytics and Reasoning for CITY(STAR-CITY), SCOnt (Smart City Ontology),
and CityPulse
Agriculture and Food supply [3] is also needed a dictionary o track entities in
system and to manage information from seeding to selling of crops. It helps to
manage data of farm items and to keep transaction easy. Smart City, Smart Home,
Need and Relevance of Common Vocabularies and Ontologies …
145
And Smart Weather [4] have dependency among each other for various tasks and
management.
Health Sector [5] shows how LOV4IoT will help to manage complex distributed
data having different type of media contents but having same vocabulary can be easily
manages. the data access will be easy due to inter linking from different departments.
Transportation and Logistics [6] will also become easy to manage in real time.
Time required to share information will get reduced and different representations of
outcome can be done easily.
7 Background
Gyrard et al. [2] projected a semantic web search application for IoT based cities with
case study of three procedure cases FIESTA-IOT EU, Machine-to-Machine Measurement (M3), and VITAL EU scheme. This project is combination of web based data
from IoT to Semantic based but not suitable for real time interoperability practically.
Kamilaris et al. [3] proposed Agri-IoT, an IoTbased smart farming applications over
web which supports big data analysis, event detection, interoperability, online information and linked datasets accessible to end uses always. But it doesn’t standardized
data specifically used for agriculture and semantic web axioms. Noura et al. [4]
created a corpus and discussed application of semantic web in Smart City, Smart
Home, and Smart Weather. It created Knowledge Extraction for the Web of Things
(KE4WoT) a set of ontologies based on some specific domain. It is efficient one if
domain to which word belong is considered else it put word in unrelated category
or discard it. This causes outlier data which can be useful in ignored data category.
Gyrard et al. [1] suggest some techniques to make ontologies more effective with
combined ontology sets for IoT and Smart city LOV, READY4SmartCities, Open
Sensing City (OSC), in addition to LOV4IoT. Gyrard et al. [7] raise the requirement
of interoperability of data needed for semantic web [8] and proposed a framework
Machine-to-Machine Measurement (M3). But it still lacks combination of different
domains. It is difficult to combine these frameworks together. Cross domain applications [9] can be useful to collect similar type of data from different applications
but requires domain base knowledge. It provides a set of linked open rules which
can be used generally for IoT applications having cross domain data. Bermudez-Edo
et al. [10] states that semantic techniques upsurge the complexity and processing time
which makes them unsuitable for IoT. To resolve this issue they proposed IoT-Lite for
semantic sensor networks but it lacks interoperability. Linked Open Vocabularies for
IoT (LOV4IoT) [11] overcome the challenges of classification, re-engineering and
designing of interoperability. This shows evolved better results and easy to establish
technique. It is up to the mark but doesn’t consider previous established classification
and classes for ontologies. IERC Cluster Semantic (IERC AC4) [12] resolve four
interoperability issues Technical, Syntactical, Semantic and Organizational. IERC
AC4 has some shortcomings also like reuse of methodologies, validation and evaluation of ontologies, and a well-designed structure. Machine-to-Machine Measurement
146
A. Sharma and R. B. Battula
(M3) framework [13] tries to simplify domain specific issues. It highlighted uniform
nomenclature requirement to improve performance of Ontology. In perspective of
security M3 framework is not secure because it is accessible by third party easily.
IoT-O [14] an extension of M2M is proposed to use known ontologies and defines
some concepts relevant to IoT. It is based on principal of indexing of resources data;
still semantic data interoperability is unachieved.
8 IoT-Related Ontologies
The space of ontologies is divided, paying little heed to the area of intrigue. The
more extravagant a metaphysics is, the bigger zone it ranges. Subsequently,
uniqueness and crossing points with different ontologies become increasingly manysided and complex. Web of Things traverses tremendous number of areas, and extends
with the developing prominence of “shrewd gadgets”. Utilization of ontologies in
the IoT mirrors this extensiveness. There are numerous ontologies that speak to
models applicable to the IoT, including, in any case, not restricted to, gadgets, units
of estimation, information streams, information preparing, geo-location, information
provenance, PC equipment, techniques for correspondence, and so forth. Highlight
of the IoT is a brilliant gadget fit for correspondence. From this point of view, ontologies that catch the possibility of a gadget, and are entrenched in the IoT space: SSN,
SAREF, oneM2M Base Ontology, IoT-Lite, and OpenIoT. Every one of them takes an
alternate way to deal with demonstrating the IoT space at the same time, in spite of the
distinctions in conceptualization, they spread converging sections of the IoT scene.
Beneath, we talk about disparity, oppositeness and covers between these ontologies.
SSN, or “Semantic Sensor Network” [15] is a metaphysics based on sensors and
perceptions. It is an accepted augmentation of the SensorML language. SSN centers
around estimations and perceptions, ignoring equipment data about the gadget. In
particular, it portrays sensors as far as abilities, execution, use conditions, perceptions,
estimation forms, and organizations. It is profoundly measured and extendable. Truth
be told, it relies upon other ontologies in key territories (for example time, area, units)
and, for every pragmatic reason, should be stretched out before genuine execution
of a SSN-based IoT framework. SSN, detailed on head of DUL, is an ontological
reason for the IoT, as it attempts to cover any utilization of sensors in the IoT.
SAREF or “The Smart Appliances REFerence” [16], metaphysics covers the
zone of shrewd gadgets in houses, workplaces, open spots, and so forth. It doesn’t
concentrate on any mechanical or logical usage. The gadgets are described overwhelmingly by the function(s) they perform, orders they acknowledge, and states
they can be in. Those three classifications fill in as building squares of the semantic
depiction in SAREF. Components from each can be consolidated to deliver complex
depictions of multi-practical gadgets. The portrayal is supplemented by gadget benefits that offer capacities. A critical module of SAREF is the vitality and force profile
that got significant consideration, not long after its inception. SAREF utilizes WGS84
for geolocation and characterizes its own estimation units. oneM2M Base Ontology
Need and Relevance of Common Vocabularies and Ontologies …
147
(oneM2M BO) is an as of late made philosophy, with first non-draft discharge in
August 2016. It is generally little, readied for the discharge 2.0 of oneM2M details,
and planned with the expectation of giving a common ontological base, to which
different ontologies would adjust. It is like the SSN, since any solid framework
fundamentally needs to expand it before execution. It depicts gadgets in an expansive
degree, empowering (in an overall sense) particular of gadget usefulness, organizing
properties, activity and administrations. The way of thinking behind this methodology was to empower revelation of semantically differentiated assets utilizing an
insignificant arrangement of ideas. It is a base cosmology, as it doesn’t expand
some other base models, (for example, DUL or Dublin Core). Notwithstanding,
arrangements to different ontologies are known.
IoT-Lite [10] is a “launch” of the SSN, for example an immediate augmentation of a few of its modules. It is a negligible cosmology, to which a large portion
of the provisos of the SSN apply. In particular: center around sensors and perceptions, dependence on other ontologies (for example time or units ontologies), high
measured quality and extendability. The thought behind the IoT-Lite was to make
a little/light semantic model that would be less burdening (than other, increasingly
verbose and more extensive models) on gadgets that process it. Simultaneously, it
expected to cover enough ideas to be valuable. The metaphysics portrays gadgets,
articles, frameworks and administrations. The principle augmentation of the SSN,
in the IoT-Lite, lies also of actuators (to supplement sensors, as a gadget type) and
an inclusion property. It expressly utilizes ideas from a geo-location cosmology to
divide gadget inclusion and arrangement area.
OpenIoT [17] philosophy was created inside the OpenIoT venture. In any case,
here, we utilize the expression “OpenIoT” to allude to the metaphysics. It is a nearly
enormous model that reuses and joins different ontologies. Those incorporate all
modules of the SSN (the principle reason for the OpenIoT), SPITFIRE (counting
sensor systems), Event Model-F, PROV-O, WGS84, CloudDomain, SIOC, Association Ontology and others, including littler ontologies created at the DERI (at present,
Insight Center). It likewise utilizes ontologies that give premise to those listed before,
for example DUL. Other than ideas from the SSN, OpenIoT, utilizes an enormous
number of SPITFIRE ideas, for example system and sensor arrange depictions. While
some referenced ontologies are definitely not imported by the OpenIoT unequivocally, they show up in all models, documentation, what’s more, venture expectations.
Consequently, one can treat OpenIoT as a blend of portions of those. Essentially to
the SSN, OpenIoT doesn’t characterize its own area ideas and doesn’t unequivocally
import geolocation ontologies. It depends on different ontologies for that yet, rather
than the SSN, it plainly shows LinkedGeoData and WGS84 of geolocation portrayals.
It characterizes a constrained set of units of measure (for example temperature, wind
speed), however just when they were pertinent to the OpenIoT venture pilot usage.
The rich set-up of utilized ontologies implies that OpenIoT gives rich depiction
of gadgets, their functionalities, capacities, provenance, estimations, arrangements
and position, vitality, pertinent occasions, clients and numerous others. Strikingly
enough, it doesn’t unequivocally portray actuators or impelling properties/capacities.
148
A. Sharma and R. B. Battula
It tends to be seen that the expansive extent of the cosmology makes it rather entangled. This is likewise in light of the fact that, it isn’t recorded all around ok, for
example the detail level and straightforward entry of the documentation don’t coordinate the range of inclusion of ideas in the model. Besides, it isn’t plainly and
expressly modularized, notwithstanding being an expansion of the SSN.
Let us note that, while there are other IoT models of expected intrigue (such
as OGC Sensor Things, FAN FPAI, UniversAAL ontologies, IoT Ontology3, M3
Vocabulary), we won’t consider them here. This is a result of space constraint, and
the way that they have produced substantially less “general intrigue”. In any case,
we intend to remember these ontologies for resulting work.
Let us presently think about the choose ontologies [18] one next to the other.
Chosen key angles, or classifications, straightforwardly relating to the IoT; set the
first segment of Table 1. Notwithstanding, due to complexities and different ways
of thinking behind thought about ontologies, every classification should be further
researched.
Every one of considered ontologies proposes an alternate way to deal with
demonstrating the IoT space [19]. The greatest contrasts are in the subtleties.
(a) OneM2M BO proposes a little base cosmology, like upper ontologies that gives
just an insignificant set of profoundly unique elements. This takes into consideration a wide arrangement of space ontologies to be effortlessly lined up with
it. It likewise implies that the BO itself isn’t sufficient to model any solid issue
Table 1 IoT ontologies comparison (a), (b)
(a)
Sub
Domain
Thing Gadget Gadget
sending
Gadget properties and
capacities
Gadget Capacity and
vitality administration
SSN
✓
✓
✓
✓
✓
SAREF
✓
✓
✓
OneM2M ✓
BO
✓
✓
IoT-Lite
✓
✓
✓
OpenIoT
✓
✓
✓
✓
✓
✓
✓
✓
✓
(b)
Sub
Domain
Detecting and sensor
properties
Perception
SSN
✓
✓
SAREF
✓
✓
OneM2M
BO
Impelling and
actuator properties
✓
✓
✓
IoT-Lite
✓
OpenIoT
✓
✓
✓
Conditionals
Need and Relevance of Common Vocabularies and Ontologies …
(b)
(c)
(d)
(e)
149
(or arrangement) in the IoT. Moreover, it doesn’t catch a few angles that are
normal in different ontologies.
OpenIoT contrasts this way of thinking by giving a nitty gritty model to a
particular issue (for example pilot usage from the OpenIoT venture) that can
be additionally applied in an increasingly broad case, or in different arrangements. Its overwhelming utilization of outside ontologies gives high semantic
interoperability by structure.
SSN is a created model of the IoT when all is said in done, however with
solid spotlight on sensors. It depends on DUL, what’s more, is unmistakably
modularized, which makes it a decent contender for augmentations into solid
frameworks and executions. This is prove by the way that other ontologies,
assessed here, utilize it. With regards to particularity, it places itself in the center
between oneM2M BO and OpenIoT.
IoT-Lite is an expansion of chose SSN modules, essentially to incorporate actuators. Or maybe than concentrating on giving a point by point portrayal of a
delimited issue space inside the IoT, it moves toward the displaying issue from
the point of view of a usage gadget. It plans to convey a little, yet complete,
model in request to rearrange preparing of semantic data. This is additionally
its particular attributes.
SAREF is a model with a solid spotlight on its own region—of brilliant apparatuses. Despite the fact that mappings to different norms exist, SAREF was
created without any preparation to speak to a particular zone of use of the IoT.
In this region, it conveys a solid and point by point base, that is additionally
clear and simple to comprehend. Simultaneously, it is sufficiently general to be
utilized when stretched out to different areas, or arrangements.
Curiously, every one of these ontologies totally ignore equipment determinations. It appears that the “place” of a gadget in an IoT framework is significantly
more imperative to philosophy engineers than its equipment detail and coming about
abilities.
9 Semantic Web
The Semantic Web is an expansion of the World Wide Web [20] through norms set
by the World Wide Web Consortium (W3C). The objective of the Semantic Web
is to make Internet information machine-decipherable. To empower the encoding
of semantics with the information, innovations, for example, Resource Description
Framework (RDF) and Web Ontology Language (OWL)are utilized. These advancements are utilized to officially speak to metadata. For instance, metaphysics can
portray ideas, connections among elements, and classifications of things. These
implanted semantics offer critical points of interest, for example, thinking over
information and working with heterogeneous information sources.
150
A. Sharma and R. B. Battula
These principles advance regular information configurations and trade conventions on the Web, on a very basic level the RDF. As per the W3C, “The Semantic
Web gives a typical structure that permits information to be shared and reused across
application, undertaking, and network boundaries.” The Semantic Web is in this
way viewed as an integrator across various substance and data applications and
frameworks.
The term was authored by Tim Berners-Lee for a snare of information (or information web) that can be prepared by machines that is, one in which a significant part
of the importance is machine-intelligible. While its faultfinders have scrutinized its
possibility, advocates contend that applications in library and data science, industry,
science and human sciences research have just demonstrated the legitimacy of the
first concept.
Berners-Lee initially communicated his vision of the Semantic Web in 1999 as
follows:
I have a fantasy for the Web [in which computers] become fit for investigating all
the information on the Web—the substance, connections, and exchanges among individuals and PCs. A “Semantic Web”, which makes this potential, still can’t seem to
develop, yet when it does, the everyday instruments of exchange, administration and
our day by day lives will be dealt with by machines conversing with machines. The
“astute specialists” individuals have promoted for a long time will at last materialize.
9.1 Difficulties
A portion of the difficulties for the Semantic Web incorporate incomprehensibility,
dubiousness, vulnerability, irregularity, and double dealing. Robotized thinking
frameworks should manage these issues so as to convey on the guarantee of the
Semantic Web.
1. Incomprehensibility: The World Wide Web contains a large number of pages.
The SNOMED CT clinical wording philosophy alone contains 370,000 class
names, and existing innovation has not yet had the option to kill all semantically copied terms. Any computerized thinking framework should manage really
immense sources of info.
2. Unclearness: These are uncertain ideas like “youthful” or “tall”. This emerges
from the unclearness of client inquiries, of ideas spoke to by content suppliers,
of coordinating inquiry terms to supplier terms and of attempting to join diverse
information bases with covering yet quietly various ideas. Fluffy rationale is the
most well-known method for managing dubiousness.
3. Vulnerability: These are exact ideas with unsure qualities. For instance, a patient
may introduce a lot of indications that relate to various diverse particular judgments each with an alternate likelihood. Probabilistic thinking strategies are
commonly utilized to address vulnerability.
Need and Relevance of Common Vocabularies and Ontologies …
151
4. Irregularity: These are consistent logical inconsistencies that will unavoidably emerge during the improvement of huge ontologies, and when ontologies
from isolated sources are joined. Deductive thinking bombs disastrously when
confronted with irregularity, since “anything follows from a logical inconsistency”. Defeasible thinking and paraconsistent thinking are two procedures that
can be utilized to manage irregularity.
5. Trickery: This is the point at which the maker of the data is deliberately deceptive the customer of the data. Cryptography methods are right now used to
lighten this danger. By giving a way to decide the data’s honesty, including
what identifies with the personality of the substance that delivered or distributed
the data, anyway validity gives despite everything must be tended to in instances
of possible misleading.
This rundown of difficulties is illustrative as opposed to comprehensive, and it
centers around the difficulties to the “binding together rationale” and “evidence”
layers of the Semantic Web. The World Wide Web Consortium (W3C) Incubator
Group for Uncertainty Reasoning for the World Wide Web (URW3-XG) last report
protuberances these issues together under the single heading of “uncertainty” [21].
Many of the procedures referenced here will expect expansions to the Web Ontology
Language (OWL) for instance to comment on restrictive probabilities. This is a region
of dynamic research.
9.2 Principles
Normalization for Semantic Web with regards to Web 3.0 is under the consideration
of W3C.
9.3 Parts
The expression “Semantic Web” [22] is frequently utilized all the more explicitly
to allude to the organizations and advancements that empower it. The assortment,
organizing and recuperation of connected information are empowered by advances
that give a conventional depiction of ideas, terms, and connections inside a given
information area. These innovations are determined as W3C guidelines and include:
•
•
•
•
•
•
•
Resource Description Framework (RDF),an overall technique for depicting data
RDF Schema (RDFS)
Simple Knowledge Organization System (SKOS)
SPARQL, a RDF question language
Notation3 (N3), planned in view of human-coherence
N-Triples, an arrangement for putting away and transmitting information
Turtle (Terse RDF Triple Language)
152
A. Sharma and R. B. Battula
• Web Ontology Language (OWL), a group of information portrayal dialects
• Rule Interchange Format (RIF), a structure of web rule language lingos supporting
principle trade on the Web.
9.4 The Semantic Web Stack
The Semantic Web Stack [23] outlines the engineering of the Semantic Web. The
capacities and connections of the parts can be summed up as follows:
XML gives an essential punctuation to content structure inside records, yet
connects no semantics with the significance of the substance contained inside. XML
isn’t at present an important part of Semantic Web advancements as a rule, as option
sentence structures exists, for example, Turtle. Turtle is an accepted norm, yet has
not experienced a conventional normalization process.
XML Schema is a language for giving and limiting the structure and substance of
components contained inside XML records.
RDF is a basic language for communicating information models, which allude
to objects (“web assets”) and their connections. A RDF-based model can be spoken
to in an assortment of language structures, e.g., RDF/XML, N3, Turtle, and RDFa.
RDF is a key norm of the Semantic Web.
RDF Schema expands RDF and is a jargon for portraying properties and classes
of RDF-based assets, with semantics for summed up progressions of such properties
and classes.
OWL includes more jargon for portraying properties and classes: among others,
relations between classes (for example disjointness), cardinality (for example “precisely one”), fairness, more extravagant composing of properties, qualities of
properties (for example balance), and listed classes.
SPARQL is a convention and inquiry language for semantic web information
sources.
RIF is the W3C Rule Interchange Format. It’s a XML language for communicating
Web decides that PCs can execute. RIF gives different renditions, called tongues. It
incorporates a RIF Basic Logic Dialect (RIF-BLD) and RIF Production Rules Dialect
(RIF PRD).
9.5 Applications
The goal is to improve the ease of use and convenience of the Web and its interconnected assets by making Semantic Web Services, for example, Servers that uncover
existing information frameworks utilizing the RDF and SPARQL norms. Numerous
converters to RDF exist from various applications. Relational databases are a significant source. The semantic web server connects to the current framework without
influencing its activity.
Need and Relevance of Common Vocabularies and Ontologies …
153
Records “increased” with semantic data. This could be machine-reasonable data
about the human-justifiable substance of the archive, (for example, the designer,
title, depiction, and so forth.) or it could be absolutely metadata speaking to a lot
of realities, (for example, assets and administrations somewhere else on the site).
Note that anything that can be related to a Uniform Resource Identifier (URI) can
be portrayed, so the semantic web can reason about creatures, individuals, places,
thoughts, and so forth. There are four semantic explanation arranges that can be
utilized in HTML records; Microformat, RDFa, Microdata and JSON-LD. Semantic
markup is frequently produced consequently, as opposed to physically.
Basic metadata vocabularies (ontologies) and guides between vocabularies that
permit record makers to realize how to increase their archives so operators can utilize
the data in the provided metadata. Computerized specialists to perform undertakings
for clients of the semantic web utilizing this information.
Electronic administrations (frequently with operators of their own) to gracefully
data explicitly to specialists, for instance, a Trust administration that an operator
could inquire as to whether some online store has a background marked by helpless
assistance or spamming.
Such administrations could be helpful to open web indexes, or could be utilized
for information the executives inside an association. Business applications include:
• Encouraging the joining of data from blended sources
• Dissolving ambiguities in corporate phrasing
• Improving data recovery subsequently decreasing data over-burden and expanding
the refinement and accuracy of the information retrieved
• Distinguishing applicable data as for a given domain
• Giving dynamic help.
In a partnership, there is a shut gathering of clients and the administration can
uphold organization rules like the selection of explicit ontologies and utilization
of semantic explanation. Contrasted with the open Semantic Web there are lesser
prerequisites on adaptability and the data circling inside an organization can be
progressively confided when all is said in done; protection is less of an issue outside
of treatment of client information.
9.6 The Semantic Web Stack for the IoT
The Semantic Web Technologies Stack presents the center Semantic Web Technologies utilized at various levels of an IoT framework [24]. The mix of Semantic
Web innovations into IoT frameworks can be distinguished at three various levels.
The “demonstrating level” gives a typical comprehension of Things’ qualities and
abilities. It utilizes shared and basic acknowledged vocabularies and ontologies to
encourage the reconciliation of information created by various frameworks (for
example sensor ontologies). The “information preparing level” utilizes depiction
rationales and OWL semantics so as to empower thinking and deduction over the
154
A. Sharma and R. B. Battula
information. At last, the “IoT Administrations and Application” level uses particular portrayal furthermore, ontologies that empowers administration distribution,
disclosure, piece and adjustment.
9.6.1
Models and Meta-models: Information Bases
The general nature of the last help or application depends of the nature of each
included layer. In this specific circumstance, the primary layer is worried about
information readiness. Deciphering furthermore, understanding the information is the
primary essential in this process. This layer deals with the semantic coordination and
accumulation of information from an assortment of sources. Semantically explained
information can be changed and displayed by explicit necessities. The Model speaks
to the Thing and structures the Assertions Box (ABox), while the Meta-model depicts
the jargon used to portray the Thing and structures the Terminological Box (TBox).
A Knowledge Base is made by these two parts. At last, the meta-metamodel gives
the build jargon to the TBox.
9.6.2
Information Preparing
An IoT framework [18, 21] is, by its temperament, a conveyed framework and
handling its information should be possible at various levels. While the restricted
neighborhood data can give some essential translation furthermore, preparing in
its area of intrigue, further understanding on the information is acquired at more
elevated levels, when information from numerous sources are assembled, prepared
and associated. We stress two distinct methodologies for preparing this information:
(1) utilizing semantic reasoners and (2) utilizing Big Data explicit calculations (for
example AI).
(1) Reasoning and Inferences.
Rules and semantic arrangements (for example owl:equivalentClass,
owl:subClassOf, owl:sameAs) can be utilized to change and adjust the information to the pronounced ontologies. Contingent upon the expressiveness of the
ontologies, thinking motors can additionally gather affiliations and connections into
the information.
For information transmission and putting away in a Semantic Web setting, JSONLD, a W3C suggestion from 2014, gives a advantageous approach to serialize
RDF information. XML design is too accessible. Triplestores (for example Fuseki,
StarDog) are utilized to store RDF significantly increases. The inquiry language
for the Semantic Web is (SPARQL Protocol And RDF Query Language). It gives a
helpful method to cross examine different triplestores over HTTP.
(2) Big Data and AI calculations.
Need and Relevance of Common Vocabularies and Ontologies …
155
Semantic Advances are a superb decision for IoT frameworks [19] for two reasons:
(1) it permits sharing the information depiction through diagrams and ontologies and
(2) it permits information encoding in ontologies by means of portrayal rationale
develops. Anyway enormous amount of information joined with high expressive
ontologies, can limit reasoners’ exhibition in their inductions. IoT frameworks are
commonly used to screen, analyze, foresee and suggest activities. In metaphysics
based frameworks, the information is depicted from the earlier, this makes them less
adjusted to frameworks where the goal is to foresee and dissect practices of various
conditions and clients. Starting here of view, the coordination of AI calculations
with very much portrayed information could give better esteem administrations and
applications. An model utilizing both measurable learning and ontologies to separate
private client action is introduced.
9.6.3
IoT Services and Applications
Ontologies and semantic comments can likewise upgrade the portrayal of the offered
types of assistance. OWL-S gives semantic markup for web administrations. The
Semantic Sensor Perception Service (SemSOS) and the utilization of ontologies
for computerized organization are a few instances of semantic advances based
applications and administrations in IoT conditions.
10 Ontologies for Smart Cities
KM4City, an Italian national task, demonstrated a cosmology intended for amassing
static or dynamic brilliant city information. The creators reuse ontologies, for
example, OWL-time, DC terms, FOAF, WGS84, GoodRelations, and cosmology
transportation systems (OTNs). The undertaking is adaptable since they handle 81
million triples with a development of 4 million triples for each month. It gives
a connected information chart, perception and investigation device what’s more,
administration map applications abusing the collected information.
Semantic traffic examination and thinking for CITY (STARCITY), an IBM
venture, is conveyed in four keen urban communities, such as Dublin, Bologna,
Miami, and Rio de Janeiro. The venture is centered around planning ontologies to analyze and anticipate street gridlocks. Information preparing misuses six
heterogeneous sources.
(1)
(2)
(3)
(4)
(5)
(6)
Road climate conditions.
Weather data.
Dublin transport stream.
Social media takes care of.
Road works and support.
City occasions.
156
A. Sharma and R. B. Battula
Semantic Web Rule Language (SWRL) rules have been intended to characterize
rules, for example, substantial traffic stream. Celebration IoT is a H2020 European
undertaking. The Celebration IoT cosmology is intended to bring together existing
IoTrelated ontologies to structure information created by testbeds. The SmartSantander city or even keen structures are testbeds creating genuine information, which
is semantically commented on as per the philosophy.
Imperative, a FP7 European undertaking, planned a cosmology to manage heterogeneous information streams produced by gadgets inside keen urban areas. The metaphysics models sensors and their, for IoT frameworks and administrations, and for
keen city applications. Indispensable is creative since it gives a working framework
to IoT to manage administration creation, coordination, and conventions. Fundamental gives the accompanying attributes: virtualization, measured quality, norms
based (RDF and JSON-LD) and inexactly coupled, also, open-source.
CityPulse, a FP7 European undertaking, gives the SAO to bring together savvy
city datasets. SAO has been structured to address ongoing viewpoints. Brilliant city
philosophy (SCO) is a metaphysics distributed in 2015. It reuses a few ontologies,
for example, SKOS, yet it doesn’t reuse the SSN philosophy and absences of best
practices. For example, the cosmology isn’t partaken in a appropriate way.
Shrewd city SOFIA2 philosophy doesn’t expand SSN cosmology be that as
it may, reuses IoT.est philosophy. PRISMA venture planned a philosophy which
reuses WGS81, NeoGeo, and assortments ontologies. Notwithstanding, it makes
reference to neither the utilization of information created by gadgets nor the utilization of SSN cosmology. The cosmology is predominantly intended to bind together
heterogeneous information.
(1) GeoData from the geographic data framework, information on lines, and stops
of the open vehicle transport framework (REST Web administration in JSON
group).
(2) Public lighting framework for the upkeep of the city (XML document).
(3) State of the streets, walkways, signs, and markings (Microsoft SQL Server
database).
(4) Historical information on city squander assortment (Microsoft Exceed expectations document).
(5) Historical information on the urban flaw detailing administration (MySQL
Server database).
The venture gives the LODView instrument to a HTML portrayal of RDF assets
and the LODLive apparatus to peruse the RDF diagram. This paper doesn’t concentrate on the depiction of the philosophy, however presents the need of this cosmology
to give connected open information and executes Web administrations, SPARQL
endpoints, browsable highlights, and representation on head of it. Brilliant city metaphysics (SCOnt) has been planned and utilized in a semantic-based system to control
brilliant city information. In any case, the cosmology has not been shared online
which ruins interoperability of shrewd city frameworks and the reuse of the philosophy. The cosmology reuses a populace philosophy, a geo-area metaphysics and the
DBpedia metaphysics. Depictions with respect to the structure of the metaphysics
Need and Relevance of Common Vocabularies and Ontologies …
157
and semantic planning are absent. The curiosity contrasted with existing brilliant city
ventures isn’t clearly clarified. SCOnt is utilized to control keen city information in
an engineering including four layers.
(1) Data scratching layer assembles and refines information since duplication and
inadequacy of meta-data and missing qualities issues are confronted.
(2) Data adjustment layer gives cosmology displaying and semantic planning.
(3) Data the board layer stores and records information inside a NoSQL database.
Semantic Web administrations are referenced yet neither connection nor
portrayals are given or on the other hand referenced.
(4) Applications layer gives dashboards and APIs.
Smart city ontologies are routinely upgraded which thwarts semantic interoperability. More ontologies identified with keen urban areas can be found on the
LOV4IoT and OSC philosophy indexes.
10.1 Measures to Compare Smart Cities and IoT Ontologies
For examination of keen city ontologies in a lot of measures to think about smart
city ontologies which can likewise be applied to IoT ontologies. Those standards are
primarily centered around the reusability of the ontologies.
(1) Ontology objective ought to be obviously clarified. As a rule, the cosmology is
intended for a task or an application.
(2) Ontology size shows the profundity of the cosmology. Little or lightweight
ontologies would be simpler to reuse.
(3) Ontology documentation decreases the expectation to absorb information to
comprehend and incorporate the metaphysics, and support its reusability. A well
known practice is to give an on the web HTML documentation. A distribution,
deliverable or any documentation is important to clarify in detail the metaphysics
and its effect.
(4) Ontology accessibility is emphatically empowered. Metaphysics ought to be
shared on the Web to energize semantic interoperability. Philosophy planners
should make an exertion in coordinating past ontologies and staying alert of the
metaphysics constraints.
(5) Ontology prominence shows the effect of the metaphysics and its genericity
when the cosmology is utilized in different tasks.
(6) Ontology support should be accomplished. Ordinarily, at the point when the
undertakings are done, the metaphysics isn’t kept up. Be that as it may,
cosmology creators may be responsive on the off chance that they keep on
taking a shot at a similar exploration point.
(7) Ontology metadata is basically required for building programmed instruments.
158
A. Sharma and R. B. Battula
11 Future Trends and Conclusion
Ontology and Vocabulary both are essential components for IoT. The cross linking
and interoperability of data make it difficult to create a standard set difficult. It requires
timely maintenance with up gradation in rules. Security on data is one of the most
needed aspects but similar use of individuals can make it risky. The interoperability
of data is still not possible due to different vocabulary present for IoT which have
no interconnection among each other. Each platform uses its own set of Axioms and
programming languages which make integration difficult. Scrutinize the evaluation
pattern of ontology is still undefined and not considered.
References
1. Gyrard, A., Zimmermann, A., Sheth, A.: Building IoT-based applications for smart cities: how
can ontology catalogs help? IEEE Internet Things J. 5(5), 3978–3990 (2018)
2. Gyrard, A., Serrano, M.: A unified semantic engine for internet of things and smart cities: from
sensor data to end-users applications. In: 2015 IEEE International Conference on Data Science
and Data Intensive Systems, pp. 718–725. IEEE (2015)
3. Kamilaris, A., Gao, F., Prenafeta-Boldu, F. X., Ali, M.I.: Agri-IoT: A semantic framework
for internet of things-enabled smart farming applications. In: 2016 IEEE 3rd World Forum on
Internet of Things (WF-IoT), pp. 442–447. IEEE (2016)
4. Noura, M., Gyrard, A., Heil, S., Gaedke, M.: Automatic knowledge extraction to build semantic
web of things applications. IEEE Internet Things J. 6(5), 8447–8454 (2019)
5. Gyrard, A., Atemezing, G., Bonnet, C., Boudaoud, K., Serrano, M.: Reusing and unifying
background knowledge for internet of things with LOV4IoT. In: 2016 IEEE 4th International
Conference on Future Internet of Things and Cloud (FiCloud), pp. 262–269. IEEE (2016)
6. Ganzha, M., Paprzycki, M., Pawłowski, W., Szmeja, P., Wasielewska, K.: Semantic interoperability in the internet of things: an overview from the INTER-IoT perspective. J. Netw. Comput.
Appl. 81, 111–124 (2017)
7. Gyrard, A., Datta, S.K., Bonnet, C., Boudaoud, K.: Cross-domain internet of things application
development: M3 framework and evaluation. In: 2015 3rd International Conference on Future
Internet of Things and Cloud, pp. 9–16. IEEE (2015)
8. Szilagyi, I., Wira, P.: Ontologies and semantic web for the internet of things-a survey. In: IECON
2016-42nd Annual Conference of the IEEE Industrial Electronics Society, pp. 6949–6954.
IEEE (2016)
9. Gyrard, A., Bonnet, C., Boudaoud, K.: Enrich machine-to-machine data with semantic web
technologies for cross-domain applications. In: 2014 IEEE World Forum on Internet of Things
(WF-IoT), pp. 559–564. IEEE (2014)
10. Bermudez-Edo, M., Elsaleh, T., Barnaghi, P., Taylor, K.: IoT-Lite: a lightweight semantic model
for the internet of things and its use with dynamic semantics. Pers. Ubiquit. Comput. 21(3),
475–487 (2017)
11. Gyrard, A., Bonnet, C., Boudaoud, K., Serrano, M.: LOV4IoT: a second life for ontologybased domain knowledge to build semantic web of things applications. In: 2016 IEEE 4th
International Conference on Future Internet of Things and Cloud (FiCloud), pp. 254–261.
IEEE (2016)
12. Gyrard, A., Serrano, M., Atemezing, G. A.: Semantic web methodologies, best practices and
ontology engineering applied to Internet of Things. In: 2015 IEEE 2nd World Forum on Internet
of Things (WF-IoT), pp. 412–417. IEEE (2015)
Need and Relevance of Common Vocabularies and Ontologies …
159
13. Gyrard, A., Datta, S. K., Bonnet, C., Boudaoud, K. Standardizing generic cross-domain applications in Internet of Things. In: 2014 IEEE Globecom Workshops (GC Wkshps), pp. 589–594.
IEEE (2014)
14. Alaya, M.B., Medjiah, S., Monteil, T., Drira, K.: Toward semantic interoperability in oneM2M
architecture. IEEE Commun. Mag. 53(12), 35–41 (2015)
15. Compton, M., Barnaghi, P., Bermudez, L., GarcíA-Castro, R., Corcho, O., Cox, S., Graybeal,
J., Hauswirth, M., Henson, C., Herzog, A., Huang, V.: The SSN ontology of the W3C semantic
sensor network incubator group. J Web Semant. 17, 25–32 (2012)
16. Daniele, L., den Hartog, F., Roes, J.: Created in close interaction with the industry: the smart
appliances reference (SAREF) ontology. In: International Workshop Formal Ontologies Meet
Industries, pp. 100–112. Springer, Cham (2015)
17. Soldatos, J., Kefalakis, N., Hauswirth, M., Serrano, M., Calbimonte, J. P., Riahi, M., Aberer, K.,
Jayaraman, P.P., Zaslavsky, A., Žarko, I.P., Skorin-Kapov, L.: Openiot: Open source internetof-things in the cloud. In: Interoperability and open-source solutions for the internet of things,
pp. 13–25. Springer, Cham (2015)
18. Bajaj, G., Agarwal, R., Singh, P., Georgantas, N., Issarny, V.: A study of existing Ontologies
in the IoT-domain. arXiv preprint arXiv:1707.00112 (2017)
19. Lam, A.N., Haugen, Ø.: Applying semantics into service-oriented iot framework. In: 2019
IEEE 17th International Conference on Industrial Informatics (INDIN), vol. 1, pp. 206–213.
IEEE (2019)
20. Mishra, S., Jain, S.: Ontologies as a semantic model in IoT. Int. J. Comput. Appl. 42(3), 233–243
(2020)
21. Ganzha, M., Paprzycki, M., Pawłowski, W., Szmeja, P., Wasielewska, K.: Towards common
vocabulary for IoT ecosystems—preliminary considerations. In: Asian Conference on Intelligent Information and Database Systems, pp. 35–45. Springer, Cham (2017)
22. Venceslau, A., Andrade, R., Vidal, V., Nogueira, T., Pequeno, V.: IoT Semantic interoperability:
a systematic mapping study. In: International Conference on Enterprise Information Systems,
vol. 1, pp. 535–544. SciTePress (2019)
23. Rhayem, A., Mhiri, M.B.A., Gargouri, F.: Semantic web technologies for the internet of things:
systematic literature review. Internet Things 100206 (2020)
24. Zgheib, R., Conchon, E., Bastide, R.: Semantic middleware architectures for IoT healthcare
applications. In: Enhanced Living Environments, pp. 263–294. Springer, Cham (2019)