The objective of this assignment is to apply text analytics techniques on a dataset of containing public tweets scraped from X (formerly Twitter) to extract valuable insights and sentiments. Students will preprocess the text data and perform descriptive analytics (word cloud and concordance) and predictive analytics such as sentiment analysis, and topic modeling to identify key topics discussed in the reviews.

Assignment 2: Individual Coursework (50%)

You will be provided with a dataset scraped from X (Twitter). The dataset includes tweets of public(consumers) about lab-grown meat also known as cultured or cell-based meat. Cultured meat is a form of cellular agriculture where meat is produced by culturing animal cells in vitro. The data also contains meta data of users that can also be utilized to generate meaningful insights.

Using text analytics techniques, analyse the public acceptance of lab-based meat. And provide actionable insights and recommendations policy makers as well as companies dealing with such foods for improving their product/services.  

 

About the dataset

Data Context

This dataset contains 91,116 tweets about lab-grown meat scraped from X (Twitter).

Data Content

The csv file contains 12 fields as below:

User id

User name

author_followers                        

author_tweets                             

author_description                     

author_location                           

text                                                 

created_at                                    

retweets                                        

replies                                            

likes                                                

quote_count

 

Tasks:

 

Data Preprocessing

Preprocess the given dataset to clean the text data. Include steps such as:

Removing irrelevant characters, numbers, and special symbols.

Converting text to lowercase.

Tokenization and stop-word removal.

Lemmatization or stemming.

Document the preprocessing steps and rationale behind each step.

 

Word Cloud

identifying frequently mentioned keywords in the reviews.

Visualizing the most common adjectives or adverbs used to describe a product or service.

 

Concordance

 

Analyzing the usage of a specific terms in the reviews.

Understanding the context in which a particular word is employed to derive its meaning and nuances.

 

Sentiment Analysis

 

Perform sentiment analysis on the preprocessed text data to determine the sentiments expressed in the review.

Analyze the distribution of sentiments and visualize the results using appropriate charts or graphs.

Provide insights into the overall sentiment of the hotel reviews and any trends observed.

 

Topic Modeling

 

Apply topic modeling techniques (e.g., Latent Dirichlet Allocation) to identify key topics in the reviews.

Analyze and interpret the topics identified, providing a brief summary of each topic.

 

Analysis and Recommendations

 

Based on the text analytics results, provide actionable insights and recommendations for whether public would acceptance such future foods to companies of this industry as well to policy makers.

Support your recommendations with evidence from the analysis.

 

Submission Guidelines:

Your report will need to show that you understand the business (lab-based meat industry in case) and that you are able to demonstrate a good understanding of the data. You are expected to produce an action plan for companies (restaurants or lab grown meat producers) involved in such alternate meats.

 

Your key tasks are:

1.       Perform stage 1 of CRISP-DM – Business Understanding (understanding of business of alternate food industry) that will justify the analysis from a business perspective. Showing what will be involved, which of the two topics you will focus on, and why there will be significant payoff for doing the work.

 

2.       Perform stage 2 of CRISP-DM – Data Understanding for the hotels that will demonstrate knowledge and understanding of the data and will present examples of the types of analysis that can be done.

 

3.        As part of your Data Understanding phase use appropriate text analytics techniques, such as word clouds, concordance, bag of words, sentiment analysis and topic analysis to provide examples of the kinds of analysis that are possible.

 

4. Make recommendations that will deliver to the hotels/restaurants with an evidence-based rationale for a 5-stage action plan to improve Business Operations, or to suggest Marketing Operations that the hotels can use to influence change.

 

Submit a comprehensive report detailing the preprocessing steps, word cloud, concordance, sentiment analysis, topic modeling results. And include visualizations, charts to support your analysis.

 

You need to complete your report of 2500 words by Wednesday 13^th of December 2023 (UK time) and it needs to be submitted via the assignment submission link in the module Canvas site.

 

Assessment Criteria:

 

Your report will be assessed on five components as below-

 

 

Report Requirement	Percentage of Marks Awarded
Quality of Report Presentation, layout, logic, impact	20
Business Understanding	20
Data Understanding	20
Use of examples in descriptive and predictive analytics to bolster the case	20
Overall Competence of Project Proposal	20

 

The word limit for the report is 2500. There will be a strict penalty for the report going beyond the +/- 10% of recommended word limit. Please note that the references would not be counted in the word limit. 

 

IMPORTANT NOTE

Plagiarism is strictly prohibited. Ensure that all work is original and properly cited. You are also reminded that evidence of plagiarism may result in disciplinary action.    

Check below the policy on academic misconduct and make sure that the work submitted is your own: https://www.stir.ac.uk/about/professional-services/student-academic-and-corporate-services/academic-registry/academic-policy-and-practice/quality-handbook/academic-integrity-policy-and-academic-misconduct-procedure/

 

Note:

Appendix

To supplement with the information, it is advised to use an appendix section for any tables, figures, screenshots, or other materials that help to justify the ideas presented. These appendices should be mentioned in the body of the report (referring the reader to the relevant appendix) and organised in the same order as they appear mentioned in the main document.  Note that, the appendix is not a section to put text that does not fit in the body of the report because of the word limit.

Word count excludes references and appendices.

In order to make your analysis more compelling, make sure that your report cites any sources/references that strengthen your argument and show evidence of your claims. These sources can be a combination of academic journal articles, statistics, practitioners’ literature, market research reports, news, etc.

Reference list

 All the references in the report should be in the reference list and the other way round (see further information at the end of the assignment brief).

Referencing

The Stirling Management School recommend using the Harvard Stirling University Referencing Style (HSU).

The following brief information will help you to get started using HSU but you should consult the Harvard Stirling University Guide on the Library web pages (http://libguides.stir.ac.uk/Harvard-Stirling)  for more detailed guidance, additional reference types and updates.  This information is also available in the Management School Undergraduate Student Handbook which is available on Canvas.

To acknowledge a paraphrased idea put the reference information in brackets next to the idea used.

For example:

There is some evidence (Smith 1995) that these figures are incorrect.

OR

Smith (1995) has provided evidence that these figures are incorrect.

Multiple Authors: If a reference has two authors include both e.g. (Smith and Richardson 2013) but if it has more than two authors give only the first name followed by et al. e.g. (Johnston et al. 2012).

Example Reference List / Bibliography

Anderson, R.C. and Klofstad, C.A. (2012) Preference for leaders with masculine voices holds in the case of feminine leadership roles. Plos One. 7 (12), e51216. Available:

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0051216 [Accessed:30 July 2014].

Creating the Kelpies (2014) [Television Broadcast] BBC 2 Scotland, 6 May.

Gilmore, S. and Williams, S. eds. (2009) Human resource management. Oxford: Oxford University Press.

Roberts, I. (2003) Sociology and industrial relations. In: P. Ackers and A. Wilkinson eds. Understanding work and employment: industrial relations in transition. Oxford: Oxford University Press, pp. 31-42.

Scottish Government (2011) Economic strategy: transition to a low carbon economy. Scottish Government. Available: http://www.scotland.gov.uk/Topics/Economy/EconomicStrategy/LowCarbon [Accessed:28 March 2012].

The Hobbit: an unexpected journey (2013) [DVD] Directed by Peter Jackson. Los Angeles: Warner Bros. Pictures.

Information on all referencing styles can be found here: http://www.stir.ac.uk/is/student/writing/referencing/howto/