Title: Apache Spark and cluster analysis for expert finding

Authors: Mohammed Zuhair Al-Taie; Seifedine Kadry

Addresses: Universiti Technologi Malaysia UTM, 81310 Skudai, Johor, Malaysia ' Beirut Arab University, Debbieh Campus, Lebanon

Abstract: Online search has become easy, thanks to the simplicity and efficiency of information access techniques. Because search engines are still unable to give answers to questions that require deep semantic understanding and enough human experience, the use of community question answering services that allow posting questions in a variety of topics has emerged. In these portals, people who provide a large number of high-quality answers are called experts. However, because a large number of questions are posted every day, experts are not always aware of the questions that appeal to them. As a result, askers need to wait for a relatively long time before receiving answers. To address this critical issue, this study aims at finding experts in community question answering ports, which will help to allow them to see the questions that appeal to them, thus shortening the time needed for answering questions and the number of unanswered questions.

Keywords: Apache Spark; cluster; online search; expert finding; StackOverflow; data mining; information retrieval; matching; precision.

DOI: 10.1504/IJKEDM.2017.091024

International Journal of Knowledge Engineering and Data Mining, 2017 Vol.4 No.3/4, pp.297 - 319

Received: 23 Mar 2017
Accepted: 26 Dec 2017

Published online: 06 Apr 2018 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article