含代码的IT社区答案质量评价模型

Code-based IT Community Answer Quality Evaluation Model

Stack Overflow问答社区已经成为软件开发者解决开发问题的重要渠道.但社区答案多样,信息繁杂,大量问答使开发者难以找到自己问题的匹配项,导致大量时间花费在寻找最佳答案上.为了解决上述问题,提出了含代码的IT社区答案质量评价模型.该模型首先收集所有符合要求的带有源码的问题答案对,分析问题中源码和答案中源码的相似程度,同时度量代码质量,然后结合用户评论对该答案的评价,统计得分,使得答案质量得以量化.最后按分数从高到低对答案进行重新排序,使代码片段质量高、相关程度高的信息出现在前列,方便用户寻找高质量的答案.实验证明,该模型能快速有效地完成IT社区答案质量评价,对答案进行重排序,给开发者迅速定位最佳答案带来非常实用的价值.结果表明,该模型切实可行.

The Stack Overflowquestions&answers community has become an important channel for software developers to solve development problems. However,the answers of the community are diverse and the information is complex. A lot of questions and answers make it difficult to find matches to their problems for developers,result in a lot of time spent looking for the best answer. In order to solve the problem,the code-based IT community answer quality evaluation model is proposed. The model firstly collects all the required questions&answers pairs with source code,analyzes the similarity between the source code of questions and answers,and measures the quality of the code at the same time. Then,it counts the score combined with the user comments on the evaluation of the answers so that the quality of the answers can be quantified. Finally,It reorders the answers according to the score from high to lowto make the code fragment’quality can be high and the information whose relevant degree is high appears in the forefront,so that users can find the answers with high relevancy conveniently. Experimental results showthat the model can evaluate the answers quality of the IT community quickly and effectively. According to reordering the answers,It is of practical value to locate the best answer quickly for the developers. The results showthat the model is feasible.