Research on Gender Prediction of MicroBlog Users Based on Machine Learning
Accurate prediction of gender information from social media, such as Weibo, is valuable for marketing and personalization.In order to identify the gender of these users, this study tries to find out a practical method to predict it.Inquiring machine learning and related theories, this article proposed a gender prediction algorithm for Microblog user, which is based on Naive Bayesian Classifier.The project used a web-spider which called APIs provided by Weibo to crawl users tweets and gender.Then we selected feature words extracted from the content posted by Microblog users to predict their genders.Three contrast experiments were set up, and the results demonstrated that the accuracy of gender prediction can be improved effectively if the algorithm takes both the frequency and discriminability of verb frequency into consideration.It is a new appliance of text mining based on machine learning technology in the era of an unprecedented amount of user-generated data.
machine learning naive bayesian classifier microblog gender prediction
XIONG Huixiang YANG Xueping NI Zhen
School of Information Management, Central China Normal University, Wuhan 430079, China
国际会议
武汉
英文
17-21
2016-10-21（万方平台首次上网日期，不代表论文的发表时间）