Recognition And Mining Of Hidden Social Relations For Accurate Portrait Of Characters

In recent years,with the widespread popularity of smart mobile terminals,online social network has become an indispensable part of people’s daily life.People use online social networks to browse information of interest and make friends with a variety of peo-ple with similar interests,these behaviors are the portray of users’ interests,hobbies and personality traits.By collecting this information,a portrait of a particular user and his or her social relationships can be created to support applications such as character analy-sis.However,for public figures,not only they appear in online social networks,but also contain a large number of relevant materials and figures in the public news media,these constitute their hidden social relations.Through the excavation of the characters hidden social relations,more accurate portraits can be achieved.Based on this,this paper studies the identification and mining of hidden social relations for accurate portraits of people.The first step of hidden social relationship mining and identification for human target is to obtain the relevant data of the human target on online social network platforms and open media platforms.To achieve this goal,a target guidance crawler system is designed in this paper,the social media platforms and public news platform will be used as the source of data acquisition,and Luo Yonghao will be used as the target guidance object to efficiently crawl the data.At the same time,aiming at some anti-crawlingmethods of the existing website,a series of optimization was carried out on the crawlersystem,combined with automatic landing and a variety of anti-shielding measures.After data acquisition,the second step is to use the acquired data to mine and iden-tify the hidden social relations.In this part,the paper proposes a method of mining and identifying hidden social relations based on heuristic clustering.First,the acquired data is processed to screen out duplicates in public media platforms and online social media platforms.Then we use heuristic clustering method to cluster the articles which contain both the target characters and their hidden social relations in the open media platform,gets the category to which the article belongs,realize the target character hidden social relations mining and identification,enrich the portrait effect of the target character...