Research of Sentiment Analysis Based on Adaboost Algorithm

Abstract

With the rapid development of the Internet, Weibo integrates the advantages of traditional websites, forums, blogs and adds various intelligent algorithms such as interactivity and intimacy between people to make data burst in a concise form. Therefore, the need for analysis and utilization of the amount of information generated by Weibo every day is even more urgent. This article uses Weibo's official API to crawl Weibo, preprocesses the Weibo text, uses the SVM algorithm to initially screen the text, uses Naive Bayes to analyze the sentiment of Weibo, and divides Weibo into positive. The three types of negative and objective, while using the Adaboost algorithm to strengthen the naive Bayes algorithm.