Trust Analysis Of User Generated Content In Web2.0Environment

We are living in the era of the internet now. The technology is growing rapidly. As the internet has entered the Web2.0era, more ordinary users get the right to speak on the internet. The real-time and various information on the Web2.0websites make more users choose to access information primarily from Web2.0websites.However, give users the right to speak also brings a series of problems. In Web2.0websites, users can publish information without supervision, which may make the information uncertain. The uncertainty can be reflected in two aspects. The first aspect is the uncertainty of users’indentities. Users’virtual indentitiesare difficult to correspond with the real identities. The second aspect is the uncertainty of information. In Web2.0websites like microblogging, there are rumors and false information, which are difficult to monitor. And the spread of false information may cause an adverse impact.For the uncertainty of information in Web2.0websites, this paper analyzes the trust problem of user generated content. The main contents are as follows.1. Data collection. In this study, collect the data from online forums and microblogging, two representatives of Web2.0websites. According to the study needs, extract the information needed from web pages. Choose a appropriate way to organize and store the data based on the needs of subsequent analysis.2. For the case that a user may have multiple virtual accounts in Web2.0websites, this study proposes a multi-demensional similarity-based algorithm to discover users’multiple virtual identities. Using the online forum datasets, some experiments are done to test the effectiveness of the proposed algorithm. The experiment results show that the algorithm proposed in this study can ef-fectively indentify the users’multiple indentities. 3. For the case that there are rumours and false information in Web2.0websites, an algorithm is proposed to analyze the credibility of user generated contents. First, using the collected microblogging datasets, analyze the different features between normal microbloggings and false microbloggings. Then, use a variety of classification algorithms, the credibility of microbloggings is indenti-fied. The experiment results show the effectiveness of the proposed algorithm. Subsequently, on the basement of the classification method, an improved al-gorithm based on sentiment analysis is proposed. The experiment results are further promoted using the improved algorithm.