TopNet: Learning from Neural Topic Model to Generate Long Stories
| Autor: | Yang, Yazheng, Pan, Boyuan, Cai, Deng, Sun, Huan |
|---|---|
| Rok vydání: | 2021 |
| Předmět: | |
| Zdroj: | Yang, Yazheng, Boyuan Pan, Deng Cai, and Huan Sun. "TopNet: Learning from Neural Topic Model to Generate Long Stories." In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, pp. 1997-2005. 2021 |
| Druh dokumentu: | Working Paper |
| DOI: | 10.1145/3447548.3467410 |
| Popis: | Long story generation (LSG) is one of the coveted goals in natural language processing. Different from most text generation tasks, LSG requires to output a long story of rich content based on a much shorter text input, and often suffers from information sparsity. In this paper, we propose \emph{TopNet} to alleviate this problem, by leveraging the recent advances in neural topic modeling to obtain high-quality skeleton words to complement the short input. In particular, instead of directly generating a story, we first learn to map the short text input to a low-dimensional topic distribution (which is pre-assigned by a topic model). Based on this latent topic distribution, we can use the reconstruction decoder of the topic model to sample a sequence of inter-related words as a skeleton for the story. Experiments on two benchmark datasets show that our proposed framework is highly effective in skeleton word selection and significantly outperforms the state-of-the-art models in both automatic evaluation and human evaluation. Comment: KDD2021, 9 pages |
| Databáze: | arXiv |
| Externí odkaz: |
|