Matching Question and Answer Using Similarity: An Experiment with Stack Overflow

Conference proceedings article


Authors/Editors


Strategic Research Themes


Publication Details

Author listPrissadang Suta, Pornchai Mongkolnam, Chun Che Fung, Jonathan H. Chan

PublisherInstitute of Electrical and Electronics Engineers Inc.

Publication year2018

Start page51

End page54

Number of pages4

ISBN9781728119786

URLhttps://ieeexplore.ieee.org/document/8783021

LanguagesEnglish-United States (EN-US)


View on publisher site


Abstract

Community question and answer (CQA) sites mostly involve knowledge-bases feeding into their automated question answering systems. This paper focuses on Stack Overflow which is an online community for developers to share knowledge in computer programming. The proposed framework consists of composing of a paired Q&A corpus, followed by building of a document model with the use of paragraph vector in distributed representation via the doc2vec method, then similarity ranking to fetch a matched answer to a given question. The model pairs the two so as to represent the semantic relevance between the questions and answers generated by the proposed method. The initial experimental results have shown the system is able to provide answers automatically and with a performance of 50% accuracy when compared to expert opinions.


Keywords

community question answeringdocument embeddingparagraph vectorsquestion answering matching


Last updated on 2023-02-10 at 07:37