Expert Retrieval System
Online support forums like StackOverflow, StackExchange, et al serve as primary general-purpose platforms used by students to search answers for questions or clarify doubts encountered while learning a programming language. Similarly, while taking a particular course, students often talk to other peers and past students who’ve taken the course to clarify their doubts about a particular topic of the course. In both of these situations, students are essentially looking to learn from peers who’re familiar with that programming language or coursework. This Master’s capstone project, Expert Retrieval System (ERS) helps students leverage their network of friends to find experts in their network at School of Electrical Engineering and Computer Science (EECS), Oregon State University (OSU). The system provides a multi-faceted search system for locating peer experts based on programming language familiarity and/or experience with OSU courses. Evaluation of ERS was done in two phases - The first phase involved collection of 42 users’ contact details, their Github profile URL (from which the system retrieved each user’s list of known programming languages), their friends in School of EECS at OSU, courses taken by them, and endorsements of programming languages’ and courses’ knowledge of their friends. The second phase involved having 16 students use ERS to find experts of a selected programming language and experts in a selected! course. All participants of the study were able to find experts known to them faster with ERS as compared to LinkedIn, and they rated ERS as highly usable.
Major Advisor: Chris Scaffidi
Committee: Arash Termehchy
Committee: Kishore Bhamidipati
Sunday, June 28, 2020 at 1:00pm to 3:00pm