10 November 2022 in IEEE/ACM Transactions on Audio, Speech, and Language Processing
Transformer structure, stacked by a sequence of encoder and decoder network layers, achieves significant development in neural machine translation. However, vanilla Transformer mainly exploits the top-layer representation, assuming the lower layers provide trivial or redundant information and thus ignoring the bottom-layer feature that is potentially valuable. In this work, we propose the G roup- Trans former model ( GTrans ) that flexibly divides multi-layer representations of both encoder and decoder into different groups and then fuses these group features to generate target words. To corroborate the effectiveness of the proposed method, extensive experiments and analytic experiments are conducted on three bilingual translation benchmarks and three multilingual translation tasks, including the IWLST-14, IWLST-17, LDC, WMT-14, WMT-21 and OPUS-100 benchmark. Experimental and analytical results demonstrate that our model outperforms its Transformer counterparts by a consistent gain. Furthermore, it can be successfully scaled up to 60 encoder layers and 36 decoder layers.
30 January 2023 in IEEE/ACM Transactions on Audio, Speech, and Language Processing
Knowledge-grounded conversation has led to great progress in producing informative dialog responses by leveraging external knowledge. This work focuses on two affiliated knowledge grounded conversation tasks: Knowledge Selection and Response Generation . Previous work followed the paradigm of selecting the most optimal knowledge piece to guide the conversation towards generating the proper response. However, some knowledge pieces, which are not recognized as optimal, may still benefit response generation. How to effectively leverage these relevant knowledge pieces for response generation still remain a tricky issue. To address this problem, we propose KIN ET , a Knowledge Incorporation Network, which deals with the problem by boosting both the knowledge selection and the response generation. The proposed model contains a negative enhanced knowledge approximator which improves knowledge selection by enhancing the dense representation of knowledge pieces, and a curriculum knowledge sampler which improves generated responses by incorporating more relevant knowledge pieces in an easy-to-hard manner. We conduct the experiment on two datasets of knowledge-grounded conversations, the results show that the proposed model significantly outperforms state-of-the-art methods in terms of both automatic and human evaluations.