Loading...

Deep Learning for Community Question Answering

Lotfi, Fariba | 2018

590 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 51553 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Beigy, Hamid; Soleymani Baghshah, Mahdie
  7. Abstract:
  8. In recent years, The Web has become an environment for sharing the knowledge of users, which as its popular examples can refer to question and answering communities. In QA communities such as Quora, StackOverflow and Yahoo! Answers, People exchange information through online questions and answers. Thus, These communities have become valuable resources of information with the participation of users. Several issues have been raised in these communities. One of the issues raised in these communities is to automatically fnd similar questions to a question and then fnd the answers that are related to it among the available answers. Other issues raised in these communities include choosing the right answer for a question, identifying the causal relationship of two occurrences to answer why questions, and expert fnding to answer the questions.The issue of duplicate question detection in community question answering is important because it helps the user to receive an appropriate response if there is a similar one to his or her question without consuming any time. Challenges in text processing such as lexical gap, polysemy, the importance of word order and data sparsity, have created failures in solving this problem. To solve these challenges, Several approaches have been proposed such as topic-based models, translation-based models and social network based models; however, They have not been able to successfully overcome this problem. In recent years, Deep learning-based approaches have achieved signifcant improvements in solving various problems. Due to growing success of deep learning, in this project we seek to solve the issue of duplicate question detection using deep learning approaches. In this research, We frst study a bilateral multi-perspective matching model for recognizing the similarity of two sentences, and then using this structure and adding the proposed cost function to Intermediate layers of the network in order to improve the problem of duplicate question detection in the forums. In evaluation, We have compared the results of our proposed method with the results of one of the most important methods implemented on the Quora dataset with accuracy, recall, precision and f1 criterion which it has better results with these four criterions
  9. Keywords:
  10. Online Forums ; Duplicate Question Detection ; Deep Learning ; Generative Adversarial Networks ; Community Question Answering

 Digital Object List

 Bookmark

No TOC