Searching for Reputable Source Code on the Web

摘要

Looking for source code on the Web is a common practice among software developers. Previous research has shown that developers use social cues over technical cues to evaluate source code candidates. However, current source code search engines do not take full advantage of social information. We present a prototype, an extension of Sourcerer, that shows reputation information for each developer involved in the project and the developer's activity level. From this implementation, we have learned that the effectiveness of this approach depends on the amount of reputation information available on the Web, which is currently scarce. We also learned that a ranking algorithm that relies on both relevance and reputation would be beneficial. More research needs to be done to explore ranking algorithms that combine both social and technical information and also reputation information for source code and people.