Graph-Based and Personalized Content Recommendations for Software Developers

Informal learning through online resources like Stack Exchange is quite popular among the software developers to ask or seek a software-related content. Personalized content recommendation for software developers is still under-explored. With years of accumulation, Stack Exchange has collected thousand of questions and answers in many fields, particularly in software engineering. In this paper, we propose a recommendation system that extracts topics from Stack Exchange and considers the developer's topical navigation from browser search logs to recommend personalized content (i.e. Stack Exchange posts) to the developer. The results shows that our trained model predicts the topics with 92% accuracy, and it is found highly reliable and transparent in terms of recommendation.