Matching Model of News and Advertisements Based on Term Vectors and Sentiment Analysis
Hongyan Li · Journal of Shanghai University of Engineering Science · 2011
Current algorithm of matching advertisements for online news,which is based on information correlation,neglecting sentiment distinctions between news and advertisements,is easy to produce mismatch.Aiming at the defect of matching algorithm,the sentiment distinction analysis was introduced,and a new model to support a better type of matching contents and sentiment between online news and advertisements was proposed based on mature techniques of key word generation and the sentiment analysis.This model has an important significance of practical application for online news websites and advertisers.