Sentiment Analysis of Public Opinion on a Tribal Presidential Candidate in the 2022 Indian Presidential Election
Dr. Vishal Kumar Singh,
Assistant Professor,
School of Management Sciences (SMS), Varanasi, India
vishalkrsingh@fmsbhu.ac.in
Dr. Pravin Chandra Singh
Assistant professor,
MSMSR, Mats University, Raipur.
pravinchandrasingh@fmsbhu.ac.in
Divya Singh
Doctoral Fellow,
Institute of Management,
Banaras Hindu University, Varanasi, U.P
divyasingh@fmsbhu.ac.in
Aditya Keshari
Assistant Professor,
Institute of Management,
Nirma University, Ahmedabad
aditya.keshari@nirmauni.ac.in
Prof. Amit Gautam
Professor,
Institute of Management,
Banaras Hindu University, Varanasi, U.P
amitgautam@fmsbhu.ac.in
Abstract
For the 2022 Indian presidential election, this study proposes a way to monitor public sentiment on the microblogging site X (Twitter) in real time. X (Twitter) has grown in popularity as a means for individuals to express their opinions about political parties and politicians. An unusual opportunity to evaluate the correlation between expressed public sentiment and election outcomes arises whenever there is a spike in X (Twitter) activity in response to breaking news or other breaking events. Furthermore, sentiment analysis can assist in determining how these events effect public perception. While typical content analysis can take days or weeks, the technology described here analyses sentiment in all election-related X (Twitter) traffic, presenting data quickly and constantly. It offers a new perspective on the workings of the electoral process and public opinion to the general people, news outlets, lawmakers, and scholars. There will be an in-depth explanation of the many components of employing sentiment analysis to forecast outcomes, as well as the possible difficulties in estimating owing to the internet's anonymity.
Keywords: Indian President Election, Droupadi Murmu, Sentiment Analysis, Social Network Analysis, Thematic Analysis.
Introduction
A relatively new topic of text analytics, sentiment analysis seeks to identify and understand underlying emotional patterns in text. This novel approach to analysis has found widespread use in CRM, especially for the purpose of complaint management (Chandak, 2022).
More businesses are embracing this technology and using it to support their marketing initiatives as interest in it grows. Due of sample bias, sentiment analysis utilising X (Twitter) has remained incredibly challenging to handle(Auxier & Anderson, 2021).
Social media gave candidates a chance to connect with their supporters and voters and express their ideas and thoughts on specific issues, and vice versa(Jain & Kumar, 2017).
Several sentiment analysis techniques were used in this work to investigate the correlation between X (Twitter) sentiment trends and election results (Markovikj et al., 2013; Sherman et al., 2015).
Political parties are becoming more divided and contentious as a result of the most recent election for president of India(Bose et al., 2019). Publicity campaigns were aided by the dynamics that inspired political allegiance and dedication(Sylwester & Purver, 2015). Numerous instances of contemporary political frenzy may be found, including the most recent political actions that took place before the election(Chaudhry et al., 2021). Additionally, the rise of social media, particularly X (Twitter), has given campaign plans a new dimension because of its capacity to simultaneously reach a huge audience and convey the voice of the common person(Kaur & Sasahara, 2016). Political party affiliation and election results have both been significantly influenced by the growing usage of social media(Chauhan et al., 2021; Yadav et al., 2022). Generally, three main goals should be accomplished with the use of X (Twitter) data.
The primary objective is to assess the degree of political election-related knowledge in a conservative country that is also highly-connected. The second objective is to develop a system that can combine online data with news about current political events so that we can evaluate the veracity of the information.
The final goal is to check the attitude of population toward the president election candidates who is currently in office using the knowledge acquired. Further, International involvement during election has seen significantly which reveal the Indian reputation in International markets (Singh & Gautam, 2022).
Background to the India Presidential Election 2022
To choose a new leader for India in 2022, as Ram Nath Kovind's term as president was about to end, the nation held its sixteenth presidential election. A total of 99.12 percent of eligible voters cast ballots in the July 18, 2022, election (Election Commission of India, 2022).
The BJP's Droupadi Murmu defeated the opposition's Yashwant Sinha in the election by a margin of 296,626 votes. Being the fifteenth leader of India, Murmu became the country's second female president and its first tribal president. Not only is she the first president to be born in India after the country's independence, but she also happens to be the youngest (Election Commission of India, 2022).
In a roundabout way, the electoral college that chooses India's president consists of elected representatives from each of the 28 states' legislatures as well as the union territories of Jammu and Kashmir, Delhi, and Puducherry.
As of 2022, there are 4,033 MLAs and 776 MPs in the electoral category and currently 90 MLAs dissolved with Jammu and Kashmir Legislative Assembly already excluded from the MLAs counting.(Election Commission of India, 2022).
The electoral committee members are each given a different number of votes by the election commission, ensuring that each states and territories should have equal opportunity of voting power corresponding to their population ratio and that the overall weight of MPs and MLAs is about equal. A majority of 543,216 votes must be cast out of the total 1,086,431 votes that the electoral college members are eligible to cast(Election Commission of India, 2022).
Both the president and vice president of India need to be Indian nationals and have reached the age of 35, according to the Indian Constitution (Article 58). The presidential candidate cannot have any paid positions inside the Indian government and must meet the same electoral requirements as a member of the Lok Sabha.
If a presidential contender requests nomination by a political party, that party will then hold an election (like a primary) to choose its nominee based on the candidate's perceived qualifications. Voters choose a group of party delegates who have pledged allegiance to a certain candidate in the primary elections; this is an example of an indirect election.
The official nomination of a candidate to run on the party's behalf is then made by the delegates. In the main election in July, voters choose a group of Electoral College electors who then choose the president and vice president. This is also a form of indirect election.
Literature Review
Researchers from fields including sociology, marketing, and computer science are now more interested in X (Twitter) overall as a result of its rise. Numerous articles in this field, particularly in marketing, have been published.
A subset of the several study groups out there looks at how social media affects business (Honey & Herring, 2009).
According to research, there are considerable variations in X (Twitter) usage and intensity. X (Twitter)s were used for anything from chats to word-of-mouth advertising(Honey & Herring, 2009; Jansen et al., 2009).
So far, research has mostly focused on X's (Twitter) generic nature, which serves a purpose but lacks the specific expertise to evaluate political concerns (Tumasjan et al., 2010).
Nearly three-quarters of Americans use social media, making it a very popular way to communicate in the current world(Auxier & Anderson, 2021).
X (Twitter) is a popular platform because it allows users to post and reply to messages called "tweets." It is a website for networking and microblogging.
There were slightly about 200 million daily active users by the year 2022 (Countries with Most X (Twitter) Users 2022, Statista, 2022). Politicians, sportsmen, and other celebrities frequently utilise X (Twitter).
New methods for doing psychological research were inspired by the enormous popularity of social media platforms. Use of social media is a sign of personality traits(Markovikj et al., 2013). X (Twitter) in particular has been used in social media research to track dynamic shifts in views, forecast political ideology, and gauge the use of moral notions(Kaur & Sasahara, 2016; Sherman et al., 2015).
By providing a cross-platform survey of how people feel about each 2020 presidential candidate's tweets, the present study builds on earlier studies. We also looked to see how frequently each moral tenet was mentioned while discussing either candidate (Blanchflower & Graham, 2021).
Web forums, blogs, and X (Twitter)s have been the subject of extensive research and discussion as an alternate arena for political discourse. Some studies have praised the quality of the most well-known political blogs, while others have questioned their ability to gather and disseminate information (Huszár et al., 2022). The populace actively engaged in politics is quite tiny, according to research, despite the fact that numerous political discussion forums and blogsto see active engagement(Gelman et al., 2021; Koop & Jansen, 2009). However, there was no further information on the general applicability of X (Twitter) in this situation (Tumasjan et al., 2010).
Recent research has mostly concentrated on the effects of social media on actual people in relation to issues like politics, public policy, and causes.
The research reviewed acknowledged the political landscape's non-online population's effect as being underappreciated (Farrell & Drezner, 2008). Numerous case studies have revealed the effectiveness of internet information as a predictor of election victory (Tumasjan et al., 2010; Williams & Gulati, 2008).
Another crucial method for locating, analysing, organising, and summarising the important themes that shed light on people's general moods is thematic analysis (Sharma & Gupta, 2021). Thematic analysis can aid with reliable and beneficial insights (Nowell et al., 2017). Interview witter data was subjected to thematic investigations, which are intended to pinpoint motifs across the events(Ahmed & Lugovic, 2018; Keshari & Gautam, 2022).
Research Framework
Using the hashtags #PresidentElection, #DraupadiMurmu data from X (Twitter) threads was collected to analyse investor attitude toward Presidential Election and Draopadi Murmu. Tweets were gathered with the use of NCapture, an NVivo addon for Google Chrome. The supported file type *.ncvx is used to import the collected database into the NVivo programme. After applying the filters below using advanced search, a total of 19,742 tweets were gathered. Word frequency analysis (WFA), which emphasises the most common terms in tweets, was used to do text analysis prior to sentiment analysis. Major hashtags that are connected to one another were the subject of a cluster analysis, which highlights how significant these hashtags are.
The NVivo software's automatic coding was used to do sentiment analysis. With the use of X (Twitter) expressions, the programme in this search for the person's viewpoint. Utilizing NVivo's automated coding tool, which highlights the principal topics that the tweets are centred upon, it was possible to determine the study's main themes.
As an add-in for Microsoft Excel 2019, NodeXL provides a versatile toolset for discovering, exploring, and getting an overview of networks. The NodeXL tool depicts networks using Microsoft Edge lists or node pairs.
One of the entities in the network is represented by each vertex. The relationship between any two vertices is represented by each edge or link connecting them. A node can be either alive or dead. The interactions, connections, or linkages that exist between these nodes constitute ties.
According to Bozkurt, (2017), the building blocks of any network are the nodes and the connections between them.
This link may go in any way; there are bidirectional connections and one-way ones (Hansen et al., 2005).
The same hashtags—#PresidentElection, #DraupadiMurmu—are used to download data from X (Twitter)with the purpose of charting the global social interactions of Indian entrepreneurs. Global tweets are included in this. Indian inhabitants communicate with the rest of the globe through networks, which are analysed using the programme NodeXL.
Objectives
The objectives are to:
Data and Methodology
From the 21st of June 2022 until the 27th of July 2022, when the campaign officially began, we gathered data from X (Twitter). With the use of the google API, the data was obtained from the X (Twitter) search engine. With a total of 19,742 tweets gathered, the data is based on the names of the Candidates. Repeated tweeting by distinct individuals is eliminated, and additional de-duplications across the various searches are carried out. To guarantee that an accurate and pure sample of tweets may be utilised for analysis, this is being done.
So, back on June 21, 2022, Yashwant Sinha, who used to lead the AITC, was chosen as the joint candidate for the presidential election by the UPA and some other opposition parties. It was quite a strong selection! On the same day, the NDA picked Droupadi Murmu as its presidential candidate.
On June 9, 2022, the Election Commission of India released the schedule for the presidential election, following the rules set out in subsection (1) of Section (4) of the Presidential and Vice-Presidential Elections Act of 1952.
Electoral notification issued by the election commission on 15 June 2022 and the deadline for submissions of candidature was June 29, 2022. Further, 18th July, 2022 declared.
We will also neglect some of the trickier elements of tweeting in this research and concentrate on the mood in each tweet (Tumasjan et al., 2010). Social media platforms now reflect popular opinion and perception of current events (Rodrigues et al., 2017).
Microblogging services such as Facebook and X (Twitter) have grown in importance as a place where individuals may express their views on current events because of how easy and accessible they are to use (Pak & Paroubek, 2010). One of the most popular social media platforms for sharing information and expressing ideas is X (Twitter).
Figure 1: Data retrieval mechanism
Source: The Author
Data Analysis and Findings
There are several study outputs obtain after data analysis. Figure 2 represent the location in India from where tweets have been posted. As New Delhi is capital of country, it shows highest posted tweet frequency of the county.
Figure 2: Number of references by location
Source: Analysis Output
Figure 3 represent the variety of hashtags used during the president election campaigning. #Presidentofindia was trending at highest frequency other than hashtags.
Figure 3: Number of references by Hashtag
Source: Analysis Output
Figure 4 interestingly revealed that tweets posted during president election outside the country was also significant. The country wise tweets density shows in the Figure 4 that reflect the popularity of Indian president elections.
Figure 4: Worldwide Tweets representation
Source: Analysis Output
The most commonly used terms in the dataset are examined using word frequency analysis in NVivo. Figure 5 represent the hashtag clustered by word similarity. In Figure 6, the word cloud produced by WFA is shown in decreasing order of usage.
A total of 100 stemmed words, each with at least five letters, were picked for the word cloud in the study.
Figure 5: Hashtags Clustered by Word similarity
Source: Analysis Output
Figure 6: Word Cloud
Source: Analysis Output
Sentiment Assessment
The goal of sentiment analysis is to identify people' opinions in texts and classify them as neutral, positive, or negative. Website reviews have been the primary focus of sentiment analysis up to now (Kamyab et al., 2018).
Because tweets provide a more diverse and richer repository of opinions and sentiments on topics like the most recent movie they watched, political issues, religious beliefs, and their attitude toward investing, analyzing X (Twitter) posts is the next critical step in the field of sentiment analysis.
Consequently, we may explore different perspectives and applications by using X (Twitter) as a corpus.
Approximately 3000 pieces of data were analysed, of which 111 were extremely negative, 400 were moderately negative, 1600 were reasonably positive, and 800 were extremely positive. Data that didn't fit into the aforementioned categories were regarded as neutral. The proportion of these attitudes is displayed in Figure 7. 17% of the dataset had a negative sentiment, whereas 63% had a positive one. This shows that despite the various negative thoughts about NDAs president candidates, public opinion is more positive.
Figure 7: X (Twitter) user Sentiment
Source: Analysis Output
Thematic Evaluation
An efficient way to compare and contrast the viewpoints of the persons used in the X (Twitter) tweets is through thematic analysis of the used data. It draws attention to the parallel viewpoint, highlights user differences, and makes surprising findings. Thematic analysis is a great tool for distilling the complex data into the key components. It forces the researcher to correctly organise the data, resulting in a report that is organised and clear at the end (Ahmed et al., 2019; Dash et al., 2020)). The main themes gleaned from the datasets are depicted in Figure 8. The fact that alerts and timings are typically followed by themes connected to president elections and their campaigning. X (Twitter) users were trying to found genuine candidate for the position of Indian presidents which reflect accordingly through trending hashtags.
Figure 8: Hashtag-wise Tweet Distribution
Source: Analysis Output
Social Networks Analysis
The study of social networks among social actors that are either implicitly or explicitly connected is known as social network analysis (SNA). The universe is made up of entities (people, organisations, artefacts, nodes, and vertices), according to social network researchers. These entities are connected through relationships. SNA concentrates on relational data about what occurs between entities rather than attribute data about individuals. The patterns that appear from significant clusters of connections are of interest to network experts. For people, SNA is more about "who you know" than "what you know" or "who you are." There are 13 clusters in Figure 9, with New Delhi, North America, and the Europe being the most notable clusters. The other clusters region exhibits interaction and shares the same sentiments amongst the nations.
|
Europe |
|
New Delhi |
|
North America |
Figure 9: Social Network Analysis
Source: Analysis Output
Discussion
This study explores the population participation in presidential election 2022. Several political opinions came out through social media network and their keywords are shown in hierarchical charts (Figure 10). After the final result of election,some may see it as symbolic, even optics, but the political importance of an Adivasi woman as President of the Republic cannot be overstated. For the first time in Indian history, a tribal chief will hold the highest constitutional office. Droupadi Murmu, was chosen as India's 15th President on with more than 60% of the total vote value. Murmu, is from Odisha, India's poorest yet most resource-rich state. Her election to the top office is seen as a victory for tribal empowerment and the community's long-ignored political aspirations. While choosing a tribal woman leader as President is a step in the right way, much work has to be done to correct the injustices suffered by tribal, who continue to be subject to physical, sexual, economic, and emotional abuse.
Figure 10: Hierarchy Chart
Source: Analysis Output
Conclusion
X (Twitter) proven to be a reliable instrument for opinion research or polling, particularly when it came to forecasting the results of a political election. In the United States, the United Kingdom, Spain, France, and Indonesia itself, several studies have been done to forecast elections. The writers of this study concentrate on tweet data on the 2022 Presidential election together with the most popular keywords. The authors put up a fresh approach to forecasting the outcome of the election that isolates the pre-processing tasks of tweet counting and sentiment analysis. Using X (Twitter) API, it is simple to get candidate tweets. Despite being easier than previous approaches, this approach produced a trustworthy result since both elements significantly impacted the prediction.
Suggestion for Future Research
The qualitative data was employed in the study since it is primarily concerned with investor mood and social interaction. Future study can use a mixed method approach that combines qualitative and quantitative data to provide a more comprehensive assessment of the political sentiment in India.
References
Ahmed, W., Bath, P. A., Sbaffi, L., & Demartini, G. (2019). Novel insights into views towards H1N1 during the 2009 Pandemic: A thematic analysis of X (Twitter) data. Health Information & Libraries Journal, 36(1), 60–72.
Ahmed, W., & Lugovic, S. (2018). Social media analytics: Analysis and visualisation of news diffusion using NodeXL. Online Information Review.
Auxier, B., & Anderson, M. (2021). Social media use in 2021. Pew Research Center, 1, 1–4.
Blanchflower, D. G., & Graham, C. L. (2021). The U Shape of Happiness: A Response. Perspectives on Psychological Science, 16(6), 1435–1446. https://doi.org/10.1177/1745691620984393
Bose, R., Dey, R. K., Roy, S., & Sarddar, D. (2019). Analyzing Political Sentiment Using X (Twitter) Data. In S. C. Satapathy & A. Joshi (Eds.), Information and Communication Technology for Intelligent Systems (pp. 427–436). Springer. https://doi.org/10.1007/978-981-13-1747-7_41
Bozkurt, A. (2017). Book Review: Analyzing Social Media Networks with NodeXL-Insights from a Connected World. Contemporary Educational Technology, 8(2), 191–194.
Chandak, A. (2022). Sentiment Analysis on Tweets in the 2020 US Presidential Election. Journal of High School Science, 32492.
Chaudhry, H. N., Javed, Y., Kulsoom, F., Mehmood, Z., Khan, Z. I., Shoaib, U., & Janjua, S. H. (2021). Sentiment Analysis of before and after Elections: X (Twitter) Data of U.S. Election 2020. Electronics, 10(17), 2082. https://doi.org/10.3390/electronics10172082
Chauhan, P., Sharma, N., & Sikka, G. (2021). The emergence of social media data and sentiment analysis in election prediction. Journal of Ambient Intelligence and Humanized Computing, 12(2), 2601–2627. https://doi.org/10.1007/s12652-020-02423-y
Countries with Most X (Twitter) users 2022, Statista. (2022, July 27). https://www.statista.com/statistics/242606/number-of-active-X (Twitter)-users-in-selected-countries/
Dash, D. M., Mohanty, P. (Dr ) A. kanta, Mohanty, M. D. O., & Sahoo, M. D. S. (2020). Emerging Entrepreneurship Opportunities in Handloom Sector with Special Reference to Odisha. Psychology and Education Journal, 57(9), 783–791. https://doi.org/10.17762/pae.v57i9.348
Election Commission of India. (2022, July 22). Press Note, Election Comission of India. Election Commission of India. https://eci.gov.in/files/file/14336-press-note/
Farrell, H., & Drezner, D. W. (2008). The power and politics of blogs. Public Choice, 134(1), 15–30.
Gelman, J., Wilson, S. L., & Sanhueza Petrarca, C. (2021). Mixing messages: How candidates vary in their use of X (Twitter). Journal of Information Technology & Politics, 18(1), 101–115. https://doi.org/10.1080/19331681.2020.1814929
Hansen, P. R., Lunde, A., & Nason, J. M. (2005). Model confidence sets for forecasting models. 1–36.
Honey, C., & Herring, S. C. (2009). Beyond microblogging: Conversation and collaboration via X (Twitter). 2009 42nd Hawaii International Conference on System Sciences, 1–10.
Huszár, F., Ktena, S. I., O’Brien, C., Belli, L., Schlaikjer, A., & Hardt, M. (2022). Algorithmic amplification of politics on X (Twitter). Proceedings of the National Academy of Sciences, 119(1), e2025334119. https://doi.org/10.1073/pnas.2025334119
Jain, V. K., & Kumar, S. (2017). Towards Prediction of Election Outcomes Using Social Media. International Journal of Intelligent Systems and Applications, 9(12), 20–28. https://doi.org/10.5815/ijisa.2017.12.03
Jansen, B. J., Zhang, M., Sobel, K., & Chowdury, A. (2009). X (Twitter) power: Tweets as electronic word of mouth. Journal of the American Society for Information Science and Technology, 60(11), 2169–2188.
Kamyab, M., Tao, R., Mohammadi, M. H., & Rasool, A. (2018). Sentiment analysis on X (Twitter): A text mining approach to the Afghanistan status reviews. Proceedings of the 2018 International Conference on Artificial Intelligence and Virtual Reality, 14–19.
Kaur, R., & Sasahara, K. (2016). Quantifying moral foundations from various topics on X (Twitter) conversations. 2016 IEEE International Conference on Big Data (Big Data), 2505–2512.
Keshari, A., & Gautam, A. (2022). Global Investing Sentiment and Social Interaction During COVID Outbreak. Review of Professional Management- A Journal of New Delhi Institute of Management, 20, 59–75. https://doi.org/10.1177/09728686221099284
Koop, R., & Jansen, H. J. (2009). Political blogs and blogrolls in Canada: Forums for democratic deliberation? Social Science Computer Review, 27(2), 155–173.
Markovikj, D., Gievska, S., Kosinski, M., & Stillwell, D. (2013). Mining Facebook Data for Predictive Personality Modeling. Proceedings of the International AAAI Conference on Web and Social Media, 7(2), 23–26.
Nowell, L. S., Norris, J. M., White, D. E., & Moules, N. J. (2017). Thematic analysis: Striving to meet the trustworthiness criteria. International Journal of Qualitative Methods, 16(1), 1609406917733847.
Pak, A., & Paroubek, P. (2010, May). X (Twitter) as a Corpus for Sentiment Analysis and Opinion Mining. Proceedings of the Seventh International Conference on Language Resources and Evaluation (LREC’10). LREC 2010, Valletta, Malta. http://www.lrec-conf.org/proceedings/lrec2010/pdf/385_Paper.pdf
Rodrigues, A. P., Rao, A., & Chiplunkar, N. N. (2017). Sentiment Analysis of Real Time X (Twitter) Data Using Big Data Approach. 2017 2nd International Conference on Computational Systems and Information Technology for Sustainable Solution (CSITSS), 1–6. https://doi.org/10.1109/CSITSS.2017.8447656
Sharma, R., & Gupta, S. (2021). Bharat towards Atmanirbharta: A X (Twitter) based analysis using NVIVO. Journal of Content, Community and Communication, 58–65.
Sherman, R. A., Rauthmann, J. F., Brown, N. A., Serfass, D. G., & Jones, A. B. (2015). The independent effects of personality and situations on real-time expressions of behavior and emotion. Journal of Personality and Social Psychology, 109(5), 872–888. https://doi.org/10.1037/pspp0000036
Singh, V. K., & Gautam, A. (2022). Investigating the Relationship Between Export Antecedents and Export Performance of Handloom Industry. FIIB Business Review, 23197145221082852.
Sylwester, K., & Purver, M. (2015). X (Twitter) Language Use Reflects Psychological Differences between Democrats and Republicans. PLOS ONE, 10(9), e0137422. https://doi.org/10.1371/journal.pone.0137422
Tumasjan, A., Sprenger, T., Sandner, P., & Welpe, I. (2010). Predicting elections with X (Twitter): What 140 characters reveal about political sentiment. Proceedings of the International AAAI Conference on Web and Social Media, 4(1), 178–185.
Williams, C., & Gulati, G. (2008). What is a social network worth? Facebook and vote share in the 2008 presidential primaries.
Yadav, D., Sharma, A., Ahmad, S., & Chandra, U. (2022). Political Sentiment Analysis: Case Study of Haryana Using Machine Learning. In N. Marriwala, C. C. Tripathi, S. Jain, & D. Kumar (Eds.), Mobile Radio Communications and 5G Networks (pp. 479–499). Springer Nature. https://doi.org/10.1007/978-981-16-7018-3_36
