SoftKG: Building A Software Development Knowledge Graph Through Wikipedia Taxonomy

Jihu Wang,Xueliang Shi,Lin Cheng,Kun Zhang,Yuliang Shi
DOI: https://doi.org/10.1109/services48979.2020.00042
2020-01-01
Abstract:At present, software development is an important way to make our life more convenient and intelligent. With the development of software programming, we have accumulated a lot of expert experience and common sense of domain knowledge. How to effectively organize and reuse these high-quality knowledge has become an urgent problem because reusing high-quality expert knowledge can greatly improve the efficiency of solving problems, especially for the novices of programming. When we encounter the programming problems, the usual solutions to get the answers is to query the search engines or consult the senior developers. However, these solutions have the following limitations: 1) the information in the field of software development is relatively scattered, for example, the demanded information is distributed on different websites. The developers need to query the search engine several times to get the information that they want, which is unfriendly, especially for the novices; 2) there is no such a unified organization form for the information we retrieve, and we need to process it further to get the answers, which is inefficient. To address the above limitations, we propose to build a software development knowledge graph (SoftKG) through Wikipedia taxonomy. Specifically, we propose a framework to build SoftKG based on open source knowledge communities.
What problem does this paper attempt to address?