基于数据挖掘的专业可信回答者个性化推荐——以Stack Overflow问答社区为例

Personalized Recommendation of Professionally Credible Answerers based on Data Mining——Taking Stack Overflow Q&A Community for Example

针对问答社区中问题不能得到及时、有效解答的现状,文章以Stack Overflow问答社区为例,首先介绍了问答社区数据的采集与预处理情况;然后,通过挖掘学习者信息,得到专业可信回答者、高信誉回答者和徽章回答者三种潜在回答者;最后,实施了三种回答者推荐并对比了推荐性能。实验结果表明,与高信誉回答者推荐和徽章回答者推荐相比,考虑回答质量和专业相关性的专业可信回答者推荐具有更高的准确率和召回率,其推荐性能更优。实施基于数据挖掘的专业可信回答者个性化推荐,能有效缓解问答社区的信息过载问题,有助于建设更高效的网络学习社区环境。

Aiming at the current situation that questions in the Q&A community cannot be timely and efficiently answered, this paper firstly introduced the data collection and pre-processing situation in the Q&A community by taking Stack Overflow Q&A community for example. Then, by mining learners' information, three kinds of potential answerers, including professionally credible answerers, highly prestige answerers and badge answerers, were proposed. Finally, three kinds of answerer recommendations were carried out and corresponding recommendation performances were compared. The experiment results showed that the recommendation of professionally credible answerers who considered answer quality and professional relevance exhibited higher precision rate, recalling rate and superior recommendation performance as compared to the recommendations of highly prestige answerers and badge answerers. The personalized recommendation of professionally credible answerers based on data mining could effectively alleviate the information overload problem in the Q&A community, and help to build a more efficient network learning community environment.