Neural Sparse Topical Coding

Peng, M, Xie, Q, Zhang, Y, Wang, H, Zhang, X, Huang, J and Tian, G 2018, 'Neural Sparse Topical Coding', in Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Melbourne, Australia, 15-20 July 2018, pp. 2332-2340.


Document type: Conference Paper
Collection: Conference Papers

Title Neural Sparse Topical Coding
Author(s) Peng, M
Xie, Q
Zhang, Y
Wang, H
Zhang, X
Huang, J
Tian, G
Year 2018
Conference name 56th Annual Meeting of the Association for Computational Linguistics
Conference location Melbourne, Australia
Conference dates 15-20 July 2018
Proceedings title Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
Publisher Association for Computational Linguistics
Place of publication ACL
Start page 2332
End page 2340
Total pages 9
Abstract Topic models with sparsity enhancement have been proven to be effective at learn- ing discriminative and coherent latent top- ics of short texts, which is critical to many scientific and engineering applica- tions. However, the extensions of these models require carefully tailored graphi- cal models and re-deduced inference al- gorithms, limiting their variations and ap- plications. We propose a novel sparsity- enhanced topic model, Neural Sparse Top- ical Coding (NSTC) base on a sparsity- enhanced topic model called Sparse Top- ical Coding (STC). It focuses on replac- ing the complex inference process with the back propagation, which makes the model easy to explore extensions. Moreover, the external semantic information of words in word embeddings is incorporated to im- prove the representation of short texts. To illustrate the flexibility offered by the neu- ral network based framework, we present three extensions base on NSTC without re-deduced inference algorithms. Experi- ments on Web Snippet and 20Newsgroups datasets demonstrate that our models out- perform existing methods.
Subjects Natural Language Processing
Copyright notice © 2018 The Association for Computational Linguistics
Versions
Version Filter Type
Access Statistics: 15 Abstract Views  -  Detailed Statistics
Created: Thu, 21 Feb 2019, 12:10:00 EST by Catalyst Administrator
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us