Understanding Language Preference for Expression of Opinion and Sentiment: What do Hindi-English Speakers do on Twitter?

Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 1131–1141, Austin, Texas, November 1-5, 2016. c⃝2016 Association for Computational Linguistics Understanding Language Preference for Expression of Opinion and Sentiment: What do Hindi-English Speakers do on Twitter? Koustav Rudra IIT Kharagpur, India koustav.rudra@cse.iitkgp.ernet.in Shruti Rijhwani∗ Carnegie Mellon University, Pittsburgh, Pennsylvania srijhwan@cs.cmu.edu Rafiya Begum Microsoft Research Labs, Bangalore, India t-rafbeg@microsoft.com Kalika Bali Microsoft Research Labs, Bangalore, India kalikab@microsoft.com Monojit Choudhury Microsoft Research Labs, Bangalore, India monojitc@microsoft.com Niloy Ganguly IIT Kharagpur, India niloy@cse.iitkgp.ernet.in Abstract Linguistic research on multilingual societies has indicated that there is usually a preferred language for expression of emotion and sen- timent (Dewaele, 2010). Paucity of data has limited such studies to participant interviews and speech transcriptions from small groups of speakers. In this paper, we report a study on 430,000 unique tweets from Indian users, specifically Hindi-English bilinguals, to un- derstand the language of preference, if any, for expressing opinion and sentiment. To this end, we develop classifiers for opinion detec- tion in these languages, and further classifying opinionated tweets into positive, negative and neutral sentiments. Our study indicates that Hindi (i.e., the native language) is preferred over English for expression of negative opin- ion and swearing. As an aside, we explore some common pragmatic functions of code- switching through sentiment detection. 1 Introduction The pattern of language use in a multilingual soci- ety is a complex interplay of socio-linguistic, dis- cursive and pragmatic factors. Sometimes speakers have a preference for a particular language for cer- tain conversational and discourse settings; on other occasions, there is fluid alteration between two or more languages in a single conversation, also known as Code-switching (CS) or Code-mixing1. Under- ∗* This work was done when the author was a Research Fel- low at Microsoft Research Lab India. 1Although some linguists differentiate between Code- switching and Code-mixing, this paper will use the two terms interchangeably. standing and characterizing language preference in multilingual societies has been the subject matter of linguistic inquiry for over half a century (see Milroy and Muysken (1995) for an overview). Conversational phenomena such as CS were ob- served only in speech and therefore, all previous studies are based on data collected from a small set of speakers or from interviews. With the grow- ing popularity of social media, we now have an abundance of conversation-like data that exhibit CS and other speech phenomena, hitherto unseen in text (Bali et al., 2014). Leveraging such data from Twitter, we conduct a large-scale study on language preference, if any, for the expression of opinion and sentiment by Hindi-English (Hi-En) bilinguals. We first build a corpus of 430,000 unique India- specific tweets across four domains (sports, enter- tainment, politics and current events) and automati- cally classify the tweets by their language: English, Hindi and Hi-En CS. We then develop an opinion de- tector for each language class to further categorize them into opinionated and non-opinionated tweets. Sentiment detectors further classify the opinionated tweets as positive, negative or neutral. Our study …

Content truncated to 500 words