A context-based model for Sentiment Analysis in Twitter

Proceedings of COLING 2014, the 25th International Conference on Computational Linguistics: Technical Papers, pages 2345–2354, Dublin, Ireland, August 23-29 2014. A context-based model for Sentiment Analysis in Twitter Andrea Vanzo and Danilo Croce and Roberto Basili Department of Enterprise Engineering University of Roma Tor Vergata Via del Politecnico 1, 00133 Roma Italy {vanzo,croce,basili}@info.uniroma2.it Abstract Most of the recent literature on Sentiment Analysis over Twitter is tied to the idea that the senti- ment is a function of an incoming tweet. However, tweets are filtered through streams of posts, so that a wider context, e.g. a topic, is always available. In this work, the contribution of this contextual information is investigated. We modeled the polarity detection problem as a sequen- tial classification task over streams of tweets. A Markovian formulation of the Support Vector Machine discriminative model as embodied by the SVMhmm algorithm has been here employed to assign the sentiment polarity to entire sequences. The experimental evaluation proves that se- quential tagging effectively embodies evidence about the contexts and is able to reach a relative increment in detection accuracy of around 20% in F1 measure. These results are particularly interesting as the approach is flexible and does not require manually coded resources. 1 Introduction Since in the Web 2.0 users can write about their life, personal experiences, share contents about facts and ideas, Social Networks became valuable sources of opinions and sentiments. This huge amount of data is crucial in the study of the interactions and dynamics of subjectivity on the Web, largely relevant for marketing tasks. Twitter is one among these microblogging services that counts about a billion of active users and 500 million of daily messages1. However, the analysis of this huge amount of information is still challenging, as language is very informal, affected by misspelling and characterized by slang and #hashtags, i.e. special user-generated tags used to contextualize different tweets around a specific topic. Researches focused on the computational study and automatic recognition of opinions and sentiments as they are expressed in free texts. It gave rise to what is currently known as Sentiment Analysis, a set of tasks aiming to detect the subjective attitude of a writer with respect to some topic. Many Sentiment Analysis studies map sentiment detection in a Machine Learning (ML) setting (Pang and Lee, 2008), where labeled data, i.e. known examples, allow to induce the detection function from real world exam- ples. In general, sentiment detection in tweets has been generally treated as any other text classification task, as proved by most papers participating to the Sentiment Analysis in Twitter task in SemEval-2013 challenge (Nakov et al., 2013): a computational representation for an incoming instance is generated by just considering one tweet at a time. The short length of the message and the resulting semantic ambi- guity are critical limitations and make the task very complex. Let us consider the following example, in which a tweet from ColMustard cites SergGray: ColMustard : @SergGray Yes, I totally agree with you about the substitutions! …

Content truncated to 500 words