Review of Deep Learning for Short Text Sentiment Tendency Analysis

Abstract

Short text sentiment tendency analysis is one of the key research issues in the field of natural language processing. Sentiment tendency analysis is a series of methods, techniques and tools used to detect the semantics of subjective inclination contained in language, and it is the key to the deep semantic understanding of text. The randomness, high ambiguity and brevity of short text data make traditional sentiment tendency analysis tasks based on feature engineering and machine learning classification technology limited. With the wide application of deep learning technology in natural language processing, the short text sentiment tendency analysis model based on deep learning has made new breakthroughs. Through combing the relevant literature, this paper first summarizes and compares traditional methods and deep learning methods, introduces and analyzes the short text sentiment tendency analysis models based on deep learning in recent years, and elaborates the connections, differences and advantages of the models. Second, it summarizes the research hotspots and progress ideas of deep learning in short text sentiment tendency analysis, and the commonly used public datasets and evaluation indicators for sentiment tendency analysis are introduced. Finally, based on the characteristics of deep learning technology and the task difficulties, the application prospect of deep learning in the direction of short text sentiment tendency analysis is predicted.

Keywords

• sentiment tendency analysis
• short text
• deep learning