Deep-Sentiment: Sentiment Analysis Using Ensemble of CNN and Bi-LSTM\n Models

Abstract

With the popularity of social networks, and e-commerce websites, sentiment\nanalysis has become a more active area of research in the past few years. On a\nhigh level, sentiment analysis tries to understand the public opinion about a\nspecific product or topic, or trends from reviews or tweets. Sentiment analysis\nplays an important role in better understanding customer/user opinion, and also\nextracting social/political trends. There has been a lot of previous works for\nsentiment analysis, some based on hand-engineering relevant textual features,\nand others based on different neural network architectures. In this work, we\npresent a model based on an ensemble of long-short-term-memory (LSTM), and\nconvolutional neural network (CNN), one to capture the temporal information of\nthe data, and the other one to extract the local structure thereof. Through\nexperimental results, we show that using this ensemble model we can outperform\nboth individual models. We are also able to achieve a very high accuracy rate\ncompared to the previous works.\n