Discovery of users' authority ranking in a community question answering website
Mincai Lai, Chunxiuzi Liu, Changxin Li, Ziqiang Yu · 2017 4th International Conference on Information, Cybernetics and Computational Social Systems (ICCSS) · 2017
With the arrival of the era of “we media”, Community Question Answering(CQA) websites become a special “we media” platforms gradually. In this type of platforms, it has evolved as an interesting problem of discover the most authoritative answer provider in different fields because the number of authoritative answer providers is a key influential factor in a successful CQA system. Recently, there are some methods mining the authoritative users who focus on the quality of the answers provided. However, the contents of answers vary a lot and will become complex if they include images or videos, which makes the works hardly can be applied in real-time applications. To address this issue, this proposal aims to mine the most authoritative answer providers, which has attracted considerable research attentions recently. In particular, we first construct a graph with every asker or answer provider being a vertex and their relationships being edges, and then propose a novel approach to compute an appropriate weight for each answer provider based on the PageRank algorithm widely used in social network search area for ranking the users preferred answers on the top. Finally, we will rank the top authoritative answer providers according to their weights to obtain the most influence ones. We still conduct extensive experiments to sufficiently evaluate the performance of our proposal.