GitHub Users Recommendations Based on Repositories and User Profile

A recommendation system is important in today’s world because it suggests many options to the user that may be useful or appealing. These kinds of recommending systems are used in many applications and devices to focus on delivering the advertisements to a specific set of people. Even the commercial and social network websites depend on the recommendation system as it helps to recommend on basis of user profiles. In this paper, we analyzed the open-source software GitHub, which is a mandatory tool for the developers as it helps them to connect and exchange their works with fellow developers. We have predicted and analyzed the spread of followers to a particular user based on their repository details and nature of relation. This system would be beneficial for the developer to approach or work with other users in the same domain. Nowadays, GitHub is one of the fast-growing platforms with millions of users sharing their projects and by implementing the recommender system helps the user to reduce their time in searching for the similar users. This recommendation system is tested on the real-time data of an existing GitHub user by implementing different concepts of social network analysis such as centrality measure, similarity measure, link prediction, triads, and natural language processing concept stemming.