Skip to Main Content
Frequently Asked Questions
Submit an ETD
Global Search Box
Need Help?
Keyword Search
Participating Institutions
Advanced Search
School Logo
Files
File List
thesis_text_may_grad_date.pdf (342.5 KB)
ETD Abstract Container
Abstract Header
Simplifying Q&A Systems with Topic Modelling
Author Info
Kozee, Troy
ORCID® Identifier
http://orcid.org/0000-0001-5114-765X
Permalink:
http://rave.ohiolink.edu/etdc/view?acc_num=akron1514592530510281
Abstract Details
Year and Degree
2017, Master of Science, University of Akron, Computer Science.
Abstract
Abstract Question answering (Q&A) systems have made great progress in recent years, but that progress has also required enormous resources in terms of computation and storage. We attempt to reduce the computational and storage footprint of Q&A systems by using topic modeling algorithms to associate questions with relevant segments of text. We compare the performance of three different Q&A systems constructed using different strategies, including topic modeling, to answer fact-based questions about three books. We compare these systems by mean total indexing time (n=10), mean index size (n=10), mean total query time over question collection (n=10), proportion of correct answers and mean reciprocal rank (MRR). We found that an NLP-based system performed best, but at the expense of much longer run times for indexing and querying. A system based around topic modeling performed worst.
Committee
Chien-Chung Chan, PhD. (Advisor)
En Cheng, PhD. (Committee Member)
Yingcai Xiao, PhD, (Committee Member)
Pages
51 p.
Subject Headings
Computer Science
Keywords
Information Retrieval
;
question-answer
;
natural language processing
;
topic modelling
;
lucene
Recommended Citations
Refworks
EndNote
RIS
Mendeley
Citations
Kozee, T. (2017).
Simplifying Q&A Systems with Topic Modelling
[Master's thesis, University of Akron]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=akron1514592530510281
APA Style (7th edition)
Kozee, Troy.
Simplifying Q&A Systems with Topic Modelling.
2017. University of Akron, Master's thesis.
OhioLINK Electronic Theses and Dissertations Center
, http://rave.ohiolink.edu/etdc/view?acc_num=akron1514592530510281.
MLA Style (8th edition)
Kozee, Troy. "Simplifying Q&A Systems with Topic Modelling." Master's thesis, University of Akron, 2017. http://rave.ohiolink.edu/etdc/view?acc_num=akron1514592530510281
Chicago Manual of Style (17th edition)
Abstract Footer
Document number:
akron1514592530510281
Download Count:
512
Copyright Info
© 2017, all rights reserved.
This open access ETD is published by University of Akron and OhioLINK.