Publications [#349652] of Lawrence Carin

Papers Published

1. Yang, Q; Huo, Z; Shen, D; Cheng, Y; Wang, W; Wang, G; Carin, L, An end-to-end generative architecture for paraphrase generation, EMNLP-IJCNLP 2019 - 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint Conference on Natural Language Processing, Proceedings of the Conference (January, 2019), pp. 3132-3142, ISBN 9781950737901
   (last updated on 2024/12/31)

   Abstract:
   Generating high-quality paraphrases is a fundamental yet challenging natural language processing task. Despite the effectiveness of previous work based on generative models, there remain problems with exposure bias in recurrent neural networks, and often a failure to generate realistic sentences. To overcome these challenges, we propose the first end-to-end conditional generative architecture for generating paraphrases via adversarial training, which does not depend on extra linguistic information. Extensive experiments on four public datasets demonstrate the proposed method achieves state-of-the-art results, outperforming previous generative architectures on both automatic metrics (BLEU, METEOR, and TER) and human evaluations.