Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd Logo
Upload

Welcome to Scribd!

  • 57 views

    Social Media Analytics Assignment

    Uploaded by

    Varun Sinha
    This document provides a sentiment analysis of customer reviews for Starbucks restaurants in a tier 3 Indian city conducted by a group of students. They extracted over 1000 reviews using web scraping and performed lexicon-based sentiment analysis in Excel with an Azure Machine Learning add-in. The analysis found 492 positive reviews, 185 neutral reviews, and 323 negative reviews, though positive reviews had the highest sentiment score. Insights suggest focusing on the many negative food product reviews or working to convert neutral reviews by enhancing the menu with popular new dishes.

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd

    Social Media Analytics Assignment

    Uploaded by

    Varun Sinha
    57 views9 pages
    This document provides a sentiment analysis of customer reviews for Starbucks restaurants in a tier 3 Indian city conducted by a group of students. They extracted over 1000 reviews using web scraping and performed lexicon-based sentiment analysis in Excel with an Azure Machine Learning add-in. The analysis found 492 positive reviews, 185 neutral reviews, and 323 negative reviews, though positive reviews had the highest sentiment score. Insights suggest focusing on the many negative food product reviews or working to convert neutral reviews by enhancing the menu with popular new dishes.

    Original Title

    SOCIAL MEDIA ANALYTICS ASSIGNMENT

    Copyright

    © © All Rights Reserved

    Available Formats

    PPTX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document provides a sentiment analysis of customer reviews for Starbucks restaurants in a tier 3 Indian city conducted by a group of students. They extracted over 1000 reviews using web scraping and performed lexicon-based sentiment analysis in Excel with an Azure Machine Learning add-in. The analysis found 492 positive reviews, 185 neutral reviews, and 323 negative reviews, though positive reviews had the highest sentiment score. Insights suggest focusing on the many negative food product reviews or working to convert neutral reviews by enhancing the menu with popular new dishes.

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    57 views9 pages

    Social Media Analytics Assignment

    Uploaded by

    Varun Sinha
    This document provides a sentiment analysis of customer reviews for Starbucks restaurants in a tier 3 Indian city conducted by a group of students. They extracted over 1000 reviews using web scraping and performed lexicon-based sentiment analysis in Excel with an Azure Machine Learning add-in. The analysis found 492 positive reviews, 185 neutral reviews, and 323 negative reviews, though positive reviews had the highest sentiment score. Insights suggest focusing on the many negative food product reviews or working to convert neutral reviews by enhancing the menu with popular new dishes.

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    of 9

    SOCIAL MEDIA

    ANALYTICS
    ASSIGNMENT
    GROUP – 10
    Abhinav Kataria - 22020343004
    Aman Kumar - 22020343010
    Gaurav Nandan – 22020343035
    Sarthak Bindroo – 22020343062
    V.V Raghu Rama Sarma - 22020343073
    Varun Kumar Sinha - 22020343076
    Sentiment Analysis For Restaurants
    Starbucks

    01 02 03
    Data Extraction Sentiment Score Actionable
    using web calculation Insights for the
    scraping company
    Data Extraction
    ◦ We have taken real data from Kaggle of the year 2019-20 that has been extracted from the internet using web scarping techniques.

    ◦ Few screenshots of the customer Review for Restaurants


    Problem Statement Of Starbucks
    ◦ Our problem statement is that the coffee chain in a tier 3 city wants to be the first choice for all kinds of food lovers in that city.

    ◦ They have consistent customer flow but haven’t seen a growth in the footfall.

    ◦ They want insights to get out of this plateau/ stagnation in the region since they believe that they haven’t tapped the full potential of
    the area.

    ◦ The management wants to do a sentiment analysis of the customer reviews to get an understanding of customer needs.
    Our Solution
    ◦ We found around 1000 customer reviews and performed sentiment analysis.

    ◦ Sentiment analysis can be done using Lexicon based approach and Supervised Learning.

    ◦ Lexicon based approach


    ◦ Mark Sentiment Words / Phrases
    ◦ Apply Sentiment Shifters
    ◦ Handle But-clauses
    ◦ Aggregate Options

    ◦ Calculation of Sentiment Score

    ◦ We used Excel add-in Azure Machine Learning for Sentiment calculation.


    Our Solution
    ◦ Excel File Link –

    ◦ We found out that out of 100 reviews 492 were positive reviews, 185 were neutral and 323 negative reviews but positive reviews has
    sentiment score of 369.3430254, for neutral the sentiment score is 98.13850492 and sentiment score for negative reviews are
    80.48952143

    Sentiments Sentiment Scores


    369.343025445
    600 400 938
    492 350
    500
    300
    400
    323 250
    300 200
    185 150 98.1385049223
    200 80.4895214331
    9
    100 336
    100
    50
    0 0
    negative neutral positive negative neutral positive
    Insights
    ◦ Findings
    ◦ There are only 182 Neutral reviews and 323 negative reviews.
    ◦ Sentiment Score for ‘Negative’ is 80.48 & for ‘Neutral’ is 98.13.

    ◦ Insights – After mulling upon our main findings we made two strategies to choose from –

    ◦ Aggressive Strategy - Here we will be strictly focus on the negative reviews

    ◦ Conservative Strategy - Here we focus on the neutral reviews and work to convert neutral reviews to positive reviews.
    Insights
    ◦ For Aggressive Approach

    ◦ Negative reviews
    ◦ Place based negative reviews – 43
    ◦ Taste based negative reviews – 9
    ◦ Food product based negative review - 271
    ◦ So we would be working to tie up with good food suppliers

    ◦ For Conservative Approach

    ◦ Our Negative sentiment Score is very less which indicates less emphatic negative reviews but a huge contribution of the total
    sentiment score is from Neutral reviews.

    ◦ We would sprinkle our menu with standout dishes that drive customers into the restaurants. So that the Neutral view slowly shifts to
    favourable reviews.
    THANKS
    GROUP - 10

    You might also like