UNIMELB at SemEval-2016 Tasks 4A and 4B: An Ensemble of Neural Networks and a Word2Vec Based Model for Sentiment Classification

This paper describes our sentiment classification system for microblog-sized documents, and documents where a topic is present. The system consists of a softvoting ensemble of a word2vec language model adapted to classification, a convolutional neural network (CNN), and a longshort term memory network (LSTM). Our main contribution consists of a way to introduce topic information into this model, by concatenating a topic embedding, consisting of the averaged word embedding for that topic, to each word embedding vector in our neural networks. When we apply our models to SemEval 2016 Task 4 subtasks A and B, we demonstrate that the ensemble performed better than any single classifier, and our method of including topic information achieves a substantial performance gain. According to results on the official test sets, our model ranked 3rd for Fin the message-only subtask A (among 34 teams) and 1st for accuracy on the topic-dependent subtask B (among 19 teams).