Design And Implementation Of Social Network Data Retrieval’s Multi-Dimensional Sorting Optimizing Algorithm

Social network is a platform within Internet for people to communicate with each other and it has become an important symbol of current network. Through social network, users can share words, pictures and videos in a timely way, which definitely increases the news broadcasting speed. Social network has deeply influenced most aspects of people’s life and because of its timeliness, sharing and interaction, it has become a main data source for people. One survey says, more than80percents of global network users are currently using social network, who are creating enormous amount of data every minute.Facing such mass data from social network, effective data retrieval according to users’expectations is crucially important. Traditional data retrieval sorting method will sort the results according to the frequencies and positions of the query keywords in each result. This is a way to represent the relevancy between the query keywords to each query result. However, social network data has some attributes different from traditional data. It has time attribute, broadcasting attribute, and other satellite information which plays an important role on data’s value. This role does not reflect in the traditional sorting algorithm. In addition, even the same query keywords, different users may have distinct query expectations. Successfully digging out users’ particular expectations is a key factor of an effective data retrieval system. Lastly, systems which provide data retrieval service may have different topics and the data related to this special topic should have more value and be returned to users in high priority.This paper firstly introduces the research background, including social network and data retrieval, and briefly describes the techniques related with data retrieval and data management. Then this paper analyzes the main data retrieval sorting algorithms and the main attributes of social network data, based on which it then lists the requirements of social network data retrieval sorting. Next, a multi-dimensional data retrieval sorting optimizing algorithm is proposed, which will optimize the original sorting result from the dimensions of data, users and application. In the end, this paper describes implementation of the above algorithm based on the data from financial microblogs in details, especially its basic structure and key modules, as well as relevant tests and analysis.