A Review About Sentiment Analysis of Short Texts Based on Machine Learning
Lisheng He · 2022
In today's Internet era, the number of netizens has grown rapidly, which has spawned a large number of comments expressing their own opinions and views.How to extract valuable emotional information from these texts and make use of them is becoming more and more important.Therefore, text sentiment analysis comes into being pregnancy.As an vital branch in the field of natural language processing.It is universally used in network public opinion monitoring and analysis, semantic network analysis, knowledge graph, content recommendation, etc.It is worthy of in-depth research by scholars.Depending on the methods of use, It can be divided into three methods, which are based on sentiment dictionary, traditional machine learning and deep learning.Among them, the classification methods based on machine learning are used commonly.Through analysizing and studying this method, summarizing the advantages and disadvantages of the specific methods, and reviewing the research results of scholars, This paper puts forward some prospects for future research directions.