A Corpus Builder: Retrieving Raw Data from GitHub for Knowledge Reuse In Requirements Elicitation

Requirement elicitation is an important task, which can lead to cost reduction in the overall software process, as it avoids failures due to lack of proper understanding about what to build. However, usually, there is a lack of time devoted to proper elicitation during software construction. We assume information from similar projects is a valuable knowledge for requirements engineers when facing a new project in the same or related domain, and its acquisition can be speeded up by knowing their main features. This information is usually located in Readme documents of GitHub. We present a tool that helps in handle this large amount of information by retrieving a corpus of Readme documents given a domain-related query. It is described, in detail, how a corpus is created and stresses the importance of having a quality corpus as base for data mining, or as input for tools of qualitative data analysis.