An Analysis of the Structure and Dynamics of Large-Scale Q/A Communities

In recent years, the World Wide Web (WWW) has transformed to a gigantic social network where people interact and collaborate in diverse online communities. By using Web 2.0 tools, people contribute content and knowledge at a rapid pace. Knowledge-intensive social networks such as Q/A communities offer a great source of expertise for crowdsourcing applications. Companies desiring to outsource human tasks to the crowd, however, demand for certain guarantees such as quality that can be expected from returned tasks. We argue that the quality of crowd-sourced tasks greatly depends on incentives and the users' dynamically evolving expertise and interests. Here we propose expertise mining techniques that are applied in online social communities. Our approach recommends users by considering contextual properties of Q/A communities such as participation degree and topic-sensitive expertise. Furthermore, we discuss prediction mechanisms to estimate answering dynamics considering a person's interest and social preferences.