Using Semantics to Search Answers for Unanswered Questions in Q&A Forums

Priyanka Singh, Elena Simperl

Abstract: The expert based question and answering forums are crowdsourced and rely on people to provide answers for questions. This paper focuses on technology based Q&A systems like StackOverflow and Reddit. These websites are popular and yet many questions remain unanswered. The Suman system uses semantic keyword search in combination with traditional text search techniques to find similar questions with answers for unanswered questions. Furthermore, the Suman system also recommends experts who can answer those questions. This helps to narrow down the long tail of unanswered questions. The Suman system utilises Semantic Web and Linked Data technologies to integrate the datasets from two websites, structure them and link them to Linked Data Cloud. It uses available tools to solve name entity disambiguation problem and expands the query term with added semantics. The Suman system was evaluated and results were analysed to show its viability.