It is of great value in social networks to recognise user's identity, which can help analyze and understand social networks, predict user's behavior, study the supervision of social networks and interaction between users. In this paper, we take user's identity in social networks as the research object. In view of the text content, multimedia content and time series content user post, we study the user's identity in social networks, which is divided into organization user and individual user, and the two user identities are defined and identified concretely. The problem in this paper is a sub-research problem in social network user analysis, and it mainly identifies the user's organization or individual identity through text content, multimedia content and time-series content published by social network users, which provided reference and help for the identification of social network user identity and further study. In social life, everyone has an identity in a specific environment relative to other people or things. Due to the different nature of users, organization users and individual users will have different characteristics in terms of social content. In social networks, user identities are essential parts of their social activities. User identities may be either explicit or implicit, and are more or less expressed in user-published content. At the same time, more user characteristics are also reflected in user postings. Therefore, the identity of social network users can be analyzed and identified through the user's content. There are many kinds of information that can reflect the user's identity in social networks, such as tag information, registration information, labels or personal statements. However, it is difficult for us to identify the authenticity or credibility of these information. The registration information and authentication information in different social platforms are also different, so it is difficult to find common information that is suitable for many social platforms. However, the content information published by users on the social platform is basically visible, such as, micro blogging, Twitter, which are relatively basic and core parts of the social platform. And, it is also relatively easy to obtain. Therefore, the method of identifying user organization-individual identity based on user content information has a certain wide adaptability. Through measuring the user's colloquialization, content (theme) complexity and the normalization of text content, and simultaneously considering the user's picture characteristic in multimedia content and the user's time series content, we propose 5 kinds of machine-operated organization-individual identity recognition (identification) methods from different angles to identify whether the user in social networks is an organization user or a individual user. Finally, in order to verify the feasibility and validity of these methods mentioned in this paper, we choose Sina micro blogging dataset to experiment and through the probability model identification method for comparative analysis. At the same time, a variety of indicators are used in the verification process to evaluate the experimental results. The experimental results show that these methods can effectively identify the organization-personal identity of users in social networks, among which the recognition results of content complexity identification method, content normalization identification method and time series content identification method are the most ideal and the accuracy rate is above 80%. © 2019, Science Press. All right reserved.
