Author: Wu, Zongheng
Title: Machine reading comprehension with deep neural networks
Advisors: Li, Wenjie (COMP)
Degree: M.Sc.
Year: 2019
Subject: Hong Kong Polytechnic University -- Dissertations
Neural networks (Computer science)
Machine learning
Artificial intelligence
Department: Department of Computing
Pages: vii, 51 pages : color illustrations
Language: English
Abstract: Recently, neural networks have been very popular in solving many tasks and shows great progress in the field of Artificial Intelligence. In this thesis, steps towards Machine reading comprehension (MRC) are examined by applying deep neural networks with attention mechanism to large-scale datasets in order to develop the human-like ability to read passages and answer corresponding questions. We explore different kinds of MRC tasks based on different public datasets. Besides we analyze the different neural network models by using three popular datasets of Stanford Question Answering Dataset (SQuAD), Microsoft Machine Reading Comprehension Dataset (MS MARCO), Baidu Machine Reading Comprehension Dataset (DuReader). We describe the methodology for machine reading comprehension. Finally, we introduce a new neural network model EaR for multi-document MRC tasks and we make a comparison of our new model with the existing state-of-art models by conducting experiments on the realistic dataset of DuReader. The thesis is closed by the future application of machine reading comprehension in the real world.
Rights: All rights reserved
Access: restricted access

Files in This Item:
File Description SizeFormat 
991022268443503411.pdfFor All Users (off-campus access for PolyU Staff & Students only)1.21 MBAdobe PDFView/Open

Copyright Undertaking

As a bona fide Library user, I declare that:

  1. I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
  2. I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
  3. I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.

By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.

Show full item record

Please use this identifier to cite or link to this item: