- 期刊
Multi-task Gated Memory Network-based Implicit Inter-sentence Relationship Recognition Analysis
摘要
Implicit Inter-sentence relationship analysis plays an important role in discourse relationship recognition. The difficulty of implicit inter-sentence relation recognition lies in the fact that the corpus of this task does not have connectives similar to explicit inter-sentence relation and lacks good features. At the same time, the scarcity of data sets and the uneven distribution of samples are also the reasons why the implicit sentence relation recognition can not be well solved. In this paper, we propose an implicit inter-sentence relationship recognition analysis method based on multi-task gated memory network. Through experiments, the method in this paper achieves good recognition performance.
延伸閱讀
- Cheng, S. Y., Guo, Z. Y., & Yin, J. (2020). Integration of Multi-granularity Information for Natural Language Inference. 電腦學刊, 31(6), 78-90. https://doi.org/10.3966/199115992020123106007
- 黃庭影（2007）。The Improvement of Encryption and Decryption on Bi-directional Association Memory based Neural Network〔碩士論文，國立臺灣師範大學〕。華藝線上圖書館。https://www.airitilibrary.com/Article/Detail?DocID=U0021-2910200810532852
- Chiang, T. C. (2018). Learning Bag-of-words Document Representation with Multi-queries Memory Networks [master's thesis, National Taiwan University]. Airiti Library. https://doi.org/10.6342/NTU201801798
- 楊知愷（2013）。Optimal Task Division and Allocation Under Hybrid Memory System Architecture〔碩士論文，中原大學〕。華藝線上圖書館。https://doi.org/10.6840/CYCU.2013.00148
- Lin, C. H. (2019). Memory-Efficient Hardware-Friendly Algorithm of Region Proposal Network for Human-Object Interaction [master's thesis, National Taiwan University]. Airiti Library. https://doi.org/10.6342/NTU201902212