UDLAP at SemEval-2016 Task 4: Sentiment Quantification Using a Graph Based Representation

Proceedings of SemEval-2016, pages 109–114, San Diego, California, June 16-17, 2016. c⃝2016 Association for Computational Linguistics UDLAP at SemEval-2016 Task 4: Sentiment Quantification Using a Graph Based Representation Esteban Castillo1, Ofelia Cervantes1, Darnes Vilari˜no2 and David B´aez1 1Universidad de las Am´ericas Puebla Department of Computer Science, Electronics and Mechatronics, Mexico {esteban.castillojz, ofelia.cervantes, david.baez}@udlap.mx 2Benem´erita Universidad Aut´onoma de Puebla Faculty of Computer Science, Mexico darnes@cs.buap.mx Abstract We present an approach for tackling the tweet quantification problem in SemEval 2016. The approach is based on the creation of a co- occurrence graph per sentiment from the train- ing dataset and a graph per topic from the test dataset with the aim of comparing each topic graph against the sentiment graphs and eval- uate the similarity between them. A heuristic is applied on those similarities to calculate the percentage of positive and negative texts. The overall result obtained for the test dataset ac- cording to the proposed task score (KL diver- gence) is 0.261, showing that the graph based representation and heuristic could be a way of quantifying the percentage of tweets that are positive and negative in a given set of texts about a topic. 1 Introduction In the past decade, new forms of communication, such as microblogging and text messaging have emerged and become ubiquitous. There is no limit to the range of information conveyed by tweets and texts. These short messages are extensively used to share opinions and sentiments that people have about their topics of interest. Working with these informal text genres presents challenges for Natural Language Processing (NLP) beyond those encoun- tered when working with more traditional text gen- res. Typically, this kind of texts are short and the language used is very informal. We can find cre- ative spelling and punctuation, slang, new words, URLs, and genre-specific terminology and abbrevi- ations that make their manipulation more challeng- ing. Representing that kind of text for automatically mining and understanding the opinions and senti- ments that people communicate inside them has very recently become an attractive research topic (Pang and Lee, 2008). In this sense, the experiments re- ported in this paper were carried out in the frame- work of the SemEval 20161 (Semantic Evaluation) which has created a series of tasks for sentiment analysis on Twitter (Nakov et al., 2016b). Among the proposed tasks we chose Task 4, subtask D which was named tweet quantification according to a two-point scale and was defined as follows: ”Given a set of tweets known to be about a given topic, estimate the distribution of the tweets across the Positive and Negative classes”. In order to solve this task we created an algorithm that builds up graphs to compare each topic against all possible sentiments for obtaining the polarity percentage of each one. The steps involved in our sentiment quan- tification process are then discussed in detail. The rest of the paper is structured as follows: in Section 2 we present some related work found in the literature with respect to the …

Content truncated to 500 words