Abstract

We aim to classify people's voting intentions by the content of their Tweets about the Scottish Independence Referendum (hereafter, IndyRef). By observing the IndyRef dataset, we find that people not only discussed the vote, but raised topics related to an independent Scotland including oil reserves, currency, nuclear weapons, and national debt. We show that the views communicated on these topics can inform us of the individuals' voting intentions (Yes vs. No). In particular, we argue that an accurate classifier can be designed by leveraging the differences in the features' usage across different topics related to voting intentions. We demonstrate improvements upon a Naive Bayesian classifier using the topics enrichment method. Our new classifier identifies the closest topic for each unseen tweet, based on those topics identified in the training data. Our experiments show that our proposed Topics-Based Naive Bayesian classifier improves accuracy by 7.8% over the classical Naive Bayesian baseline.

Highlights

  • Twitter emerged as an especially popular platform during the IndyRef held in 2014

  • We examined the networks of the 7337 users in our dataset, and identified who these users follow among the 536 public Twitter accounts corresponding to Members of the British or Scottish Parliaments

  • We vary the number of selected features F and the deployed feature selection approach for both Naive Bayes (NB) and Topics-Based Naive Bayesian (TBNB)

Read more

Summary

INTRODUCTION

We propose a technique to analyse the voting intentions of users, based on data mining and machine learning approaches. The general approach we propose could be used to understand users’ voting intentions in other major elections. We capture two months of Twitter data related to the IndyRef. To form a ground truth, we label users based upon hashtags appearing in their tweets, and we verify the reliability of this approach using the users’ followee networks. The referendum created an evolving discourse, with different topical themes (such as oil, currency, and debt ), which make the accurate classification of users’ voting intentions more challenging. The dichotomy of the term “change” in indicating voting intentions across different topics highlights the main benefit of our approach. Our approach, called Topics-Based Naive Bayesian (TBNB) demonstrates marked improvements over a classical Naive Bayes (NB) classification baseline

AND RELATED WORK
TOPICS-BASED NAIVE BAYESIAN
Findings
REFERENDUM DATA AND EXPERIMENTS
Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.