YNU-HPCC at SemEval 2017 Task 4: Using A Multi-Channel CNN-LSTM Model for Sentiment Classification

Proceedings of the 11th International Workshop on Semantic Evaluations (SemEval-2017), pages 796–801, Vancouver, Canada, August 3 - 4, 2017. c⃝2017 Association for Computational Linguistics YNU-HPCC at SemEval 2017 Task 4: Using A Multi-Channel CNN-LSTM Model for Sentiment Classification Haowei Zhang, Jin Wang, Jixian Zhang, Xuejie Zhang School of Information Science and Engineering Yunnan University Kunming, P.R. China Contact: xjzhang@ynu.edu.cn Abstract In this paper, we propose a multi-channel convolutional neural network-long short- term memory (CNN-LSTM) model that consists of two parts: multi-channel CNN and LSTM to analyze the sentiments of short English messages from Twitter. Un- like a conventional CNN, the proposed model applies a multi-channel strategy that uses several filters of different length to ex- tract active local n-gram features in differ- ent scales. This information is then sequen- tially composed using LSTM. By combin- ing both CNN and LSTM, we can consider both local information within tweets and long-distance dependency across tweets in the classification process. Officially re- leased results show that our system outper- forms the baseline algorithm. 1 Introduction Social network services (SNSs) such as Twitter, Fa- cebook, and Weibo are used daily to express thoughts, opinions, and emotions. In Twitter, 6000 short messages (tweets) are posted by users every second1. Therefore, Twitter is considered as one of the most concentrated opinion-expressing venues on the Internet. Subjective analysis of this type of user-generated content has become a vital task for politics, social networking, marketing, and adver- tising. The potential application of sentiment analysis has been the motivation behind the SemEval 2017 Task 4, which is a competition involving a series of subtasks that focus on Twitter sentiment classifica- tions. Subtask A involves message polarity classi- fication, which requires a system to classify 1 http://www.internetlivestats.com/twitter-statistics/ whether a message is of positive, negative, or neu- tral sentiment. Subtasks B and C involve topic- based message polarity classification, which re- quire a system to classify a message on two- and five-point scales toward a certain topic. Various approaches have been proposed to ana- lyze sentiment of text, and deep neural network has achieved state-of-the-art results in recent years. Proven successful text classification methods in- clude convolutional neural networks (CNN) (LeCun et al., 1990; Y. Kim, 2014; Kalchbrenner et al., 2014) and Long Short-Term Memory (LSTM) (Hochreiter et al, 1997; Tai et al., 2015). In general, CNN applies a convolutional layer to extract active local n-gram features, but lost the or- der of words. By contrast, LSTM can sequentially model texts. However, it focuses only on past in- formation and draws conclusions from the tail part of texts. It fails to capture the local response from temporal data. In this paper, we propose a multi-channel CNN- LSTM model for sentiment classification. It con- sists of two parts: multi-channel CNN, and LSTM. Unlike a conventional CNN model, we apply a multi-channel strategy that uses several filters of different length. The model is thus able to extract active n-gram features of different scales. LSTM is then applied to compose those features sequentially. By …

Content truncated to 500 words