A Dictionary-Based Convolutional Recurrent Neural Network Model for Sentiment Analysis

Natural Language Processing is an important direction in the field of computer science and artificial intelligence. It studies various theories and methods that can realize effective communication between human and computer in natural language. Sentiment analysis is a common and important task in natural language processing. It is often used to analyze and judge the sentiment types of text description. Movie reviews can provide valuable reference information for people to choose high quality movies. In order to effectively interpret movie reviews and understand the sentiment factors in movie reviews, the text proposes a deep learning method based on sentiment dictionary to realize the sentiment analysis of film reviews and to restore the real feelings of users as far as possible. The method first preprocesses the collected movie reviews data, uses the skip-gram model in the word2vec network to automatically generate word vectors, and transforms the sentiment vocabulary marked by Dalian University of Technology into word vectors. Then, the words vectors generated by text and dictionary are input into the convolutional neural network to learn sentence representation, extract local features. Finally, the Long Short-Term Memory network in recurrent neural network captures the semantic information and long-term dependencies between sentences, and inputs them into the logistic regression classifier to realize the sentiment analysis of movie reviews. The proposed method is compared with the Naive Bayesian and Support Vector Machine in the traditional machine learning method, the convolutional neural network and the recurrent neural network in the deep learning method. The experimental results show that the proposed model can extract more abundant features and achieve state-of-the-art classification effect than the baseline models.