面向专家示例的Stack Overflow本体构造和推理研究

Research on expert code example oriented ontology construction and reasoning for Stack Overflow

Stack Overflow是一个计算机领域的IT技术问答网站,为了获取问答网站中的专家示例并将其应用于API挖掘中。首先采用Scrapy爬虫框架技术获取Stack Overflow问答网站中的结构化数据,并存储在关系模式中;再使用本体建模工具Protégé构建本体,然后使用D2RQ工具实现对关系数据库的知识抽取,将关系模式转换为三元组形式的本体模型;同时,提出了一个面向专家示例的子本体抽取算法,用于从原本体中抽取出专家示例推理相关的子本体,并提出了若干条专家示例推理规则,能推导出专家所编写的代码示例。实验结果证明,从Stack Overflow本体模型中抽取的专家示例能提高API调用序列挖掘的准确率。

Stack Overflow is an IT technology Q&A website in the computer field.In order to obtain expert examples in the Q&A website and apply them to API mining,this paper firstly used the Scrapy crawler framework technology to obtain structured data in the Stack Overflow Q&A website and store it in the relational model.Then it used the ontology modeling tool Protégéto build the ontology,and then used the D2RQ tool to achieve the knowledge extraction of the relational database,and transformed the relational model into the ontology model in the form of triplets.At the same time,this article proposed a sub ontology extraction algorithm for expert examples,and used it to extract sub ontologies related to expert example reasoning from the original ontology,and proposed several expert example reasoning rules that could derive expert examples.The experimental results demonstrate that extracting expert examples from the Stack Overflow ontology model can improve the accuracy of API call sequence mining.