Research On Stock News Recommendation Method Based On Ontology

With the rapid development of the Internet,people access to information has undergone tremendous changes.At the same time,with the development of China's securities market,more and more investors has joined them.The critical problem people are confronted with is the information overload which means the volumes of stock news are overwhelming to the users.In order to meet the challenge of information overload,this paper presents a stock news recommendation algorithm to help investors find the news which they are interested.Traditional news models are based on TF-IDF,and do not consider semantic information,and will be ambiguous.In order to solve the ambiguity problem,this paper uses the ontology to optimize the TF-IDF method,and proposes the OF-IDF method to establish the vector space model for the news.This paper analyzes the user profile,combined with the behavior of investors in the system,that the interests of investors mainly from two aspects,one is the history of browsing records,the other is the concept of the user concerned(stock name).In this paper,a new ontology concept correlation algorithm is proposed,and the concept of user interest is calculated by using the algorithm.The algorithm is used to calculate the concept of user interest.And the relevance of the concept of the ontology in the news,combined with the concept of OF-IDF in the news of the weight,to quantify the relevance of news and user tags.Recommend the most relevant news with user interest tags.For the user's history browsing records,the introduction of user behavior feedback,to automatically detectthe user's real interest.Finally,the experiment proves the validity of the proposed algorithm...