Research On User Information Seeking Behavior Driven By Task In The Social Q&A Platform

Research on information seeking behavior has become the core field of user behavior research, and recently tends to focus on the specific user groups and the specific information systems. With the rise and development of the social information systems,social information seeking has heavily attracted the scholars' attention. And in different social information systems, users' information seeking behavior cannot be generalized.Now the social Q&A platform has developed into a huge and valuable knowledge base with social relations. More and more users use it to seek for information, to solve problems. However, the explosive growth of questions and answers brings much difficulty for users to find high quality contents, besides, the users' information seeking pattern is not single. So, regarding the social Q&A platform as the research object to explore the information seeking behavior becomes very necessary.In the past, related researches on the users' seeking behavior in social Q&A platform mainly involve the following aspects: seeking contents, how to seeking, result evaluation and so on. About "how to seeking",more reseaches start from one strategy or one step out of the whole seeking process. What's more, previous scholars who have been following the impact of task on information seeking behavior generally throw themselves into the web search engines or professional search systems. The findings have yet to be extended and verified in the social Q&A platform. For these reasons, this dissertation designing the task context, focuses on the entire seeking process to study the users'information seeking behavior, such as query formulation and reformulation, browsing behavior, selection behavior, seeking path.This dissertation firstly discusses the connotation, content generation and transmission mechanism of social Q&A platform, especially analyses the content organization of zhihu website, on the basis of which, combined with the relative theories,detailly defines and describes all variables of information seeking behavior, and then constructs the task framework with reference to the existing task classification. We adopts the method of experimental research, using Morae and questionnaires to collect original data, and using Morae Manager, Excel, SPSS 19.0 to conduct data analysis.Overall, experimental results are discovered and summarized from two dimensions,including users' seeking behavior characteristics and users' seeking path patterns. On the first dimension, query reformulation is fully divided into several types, and some transfer patterns of query reformulation are effectively identified by analysis of sequential query reformulations. In addition, this dissertation seeks to determine whether the different tasks are likely to significantly impact the users' query reformulation behavior, browsing behavior, and whether there is a difference in the way that users browse and select topics,questions, or other users in each task. On the second dimension, this dissertation tries to summarize three modes about seeking path, namely "query formulation/browsing","selecting questions/browsing details/other behaviors","selecting topics/browsing details/other behaviors". Finally, according to the results, six kinds of optimization strategies are put forward based on search function, platform structure and information quality.