2021/10/15 by Chandra Maddila, Maddila, Chandra, Shanbhogue, Suhas +12
Computer Science · #FOS: Computer and information sciences #Software Engineering (cs.SE) #Software Engineering Research #Software Engineering Techniques and Practices #Software System Performance and Reliability
paper · pdf · doi:10.48550/arxiv.2110.08403
openalex publication_date 2021/10/15 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28
Software development is information-dense knowledge work that requires\ncollaboration with other developers and awareness of artifacts such as work\nitems, pull requests, and files. With the speed of development increasing,\ninformation overload is a challenge for people developing and maintaining these\nsystems. Finding information and people is difficult for software engineers,\nespecially when they work in large software systems or have just recently\njoined a project. In this paper, we build a large scale data platform named\nNalanda platform, which contains two subsystems: 1. A large scale\nsocio-technical graph system, named Nalanda graph system 2. A large scale\nrecommendation system, named Nalanda index system that aims at satisfying the\ninformation needs of software developers. The Nalanda graph is an enterprise\nscale graph with data from 6,500 repositories, with 37,410,706 nodes and\n128,745,590 edges. On top of the Nalanda graph system, we built software\nanalytics applications including a newsfeed named MyNalanda, and based on\norganic growth alone, it has Daily Active Users (DAU) of 290 and Monthly Active\nUsers (MAU) of 590. A preliminary user study shows that 74% of developers and\nengineering managers surveyed are favorable toward continued use of the\nplatform for information discovery. The Nalanda index system constitutes two\nindices: artifact index and expert index. It uses the socio-technical graph\n(Nalanda graph system) to rank the results and provide better recommendations\nto software developers. A large scale quantitative evaluation shows that the\nNalanda index system provides recommendations with an accuracy of 78% for the\ntop three recommendations.\n