An Ensemble approach for Aggression Identification in English and Hindi Text

摘要

In this paper we describe our system that we develop as part of our participation in the shared task at COLING 2018 TRAC-1: Aggression Identification. The objective of this task was to predict online aggression spread through online textual post or comment. The datasets were released in two languages, one for English and the other for Hindi. For each of these languages we submitted one system. Both of our systems are based on an ensemble architecture where the individual models are based on Convoluted Neural Network (CNN) and Support Vector Machine (SVM). Our system on English facebook and social media post obtains the F1 scores of 0.5151 and 0.5099, respectively where evaluation on Hindi facebook and social media obtains Fl score of 0.5599 and 0.3790, respectively.