SeerNet at SemEval-2018 Task 1: Domain Adaptation for Affect in Tweets

Proceedings of the 12th International Workshop on Semantic Evaluation (SemEval-2018), pages 18–23 New Orleans, Louisiana, June 5–6, 2018. ©2018 Association for Computational Linguistics SeerNet at SemEval-2018 Task 1: Domain Adaptation for Affect in Tweets Venkatesh Duppada, Royal Jain, Sushant Hiray Seernet Technologies, LLC {venkatesh.duppada, royal.jain, sushant.hiray}@seernet.io Abstract The paper describes the best performing sys- tem for the SemEval-2018 Affect in Tweets (English) sub-tasks. The system focuses on the ordinal classification and regression sub-tasks for valence and emotion. For ordinal classi- fication valence is classified into 7 different classes ranging from -3 to 3 whereas emotion is classified into 4 different classes 0 to 3 sep- arately for each emotion namely anger, fear, joy and sadness. The regression sub-tasks es- timate the intensity of valence and each emo- tion. The system performs domain adaptation of 4 different models and creates an ensem- ble to give the final prediction. The proposed system achieved 1st position out of 75 teams which participated in the fore-mentioned sub- tasks. We outperform the baseline model by margins ranging from 49.2% to 76.4%, thus, pushing the state-of-the-art significantly. 1 Introduction Twitter is one of the most popular micro-blogging platforms that has attracted over 300M daily users1 with over 500M 2 tweets sent every day. Tweet data has attracted NLP researchers because of the ease of access to large data-source of peo- ple expressing themselves online. Tweets are micro-texts comprising of emoticons, hashtags as well as location data, making them feature rich for performing various kinds of analysis. Tweets provide an interesting challenge as users tend to write grammatically incorrect and use informal and slang words. In domain of natural language processing, emo- tion recognition is the task of associating words, phrases or documents with emotions from prede- fined using psychological models. The classifica- tion of emotions has mainly been researched from 1https://www.statista.com/statistics/282087/number-of- monthly-active-twitter-users/ 2http://www.internetlivestats.com/twitter-statistics/ two fundamental viewpoints. (Ekman, 1992) and (Plutchik, 2001) proposed that emotions are dis- crete with each emotion being a distinct entity. On the contrary, (Mehrabian, 1980) and (Russell, 1980) propose that emotions can be categorized into dimensional groupings. Affect in Tweets (Mohammad et al., 2018) - shared task in SemEval-2018 focuses on extract- ing affect from tweets confirming to both vari- ants of the emotion models, extracting valence (di- mensional) and emotion (discrete). Previous ver- sion of the task (Mohammad and Bravo-Marquez, 2017) focused on estimating the emotion intensity in tweets. We participated in 4 sub-tasks of Affect in Tweets, all dealing with English tweets. The sub-tasks were: EI-oc: Ordinal classification of emotion intensity of 4 different emotions (anger, joy, sadness, fear), EI-reg: to determine the inten- sity of emotions (anger, joy, sadness, fear) into a real-valued scale of 0-1, V-oc: Ordinal classifica- tion of valence into one of 7 ordinal classes [-3, 3], V-reg: determine the intensity of valence on the scale of 0-1. Prior work in extracting Valence, Arousal, Dominance (VAD) from text primarily relied on using and extending lexicons (Bestgen and Vincze, 2012) (Turney et al., 2011). Recent ad- vancements …

Content truncated to 500 words