A Document-Grounded Matching Network for Response Selection in Retrieval-Based Chatbots
Abstract
We present a document-grounded matching network (DGMN) for response selection that can power a knowledge-aware retrieval-based chatbot system. The challenges of building such a model lie in how to ground conversation contexts with background documents and how to recognize important information in the documents for matching. To overcome the challenges, DGMN fuses information in a document and a context into representations of each other, and dynamically determines if grounding is necessary and importance of different parts of the document and the context through hierarchical interaction with a response at the matching step. Empirical studies on two public data sets indicate that DGMN can significantly improve upon state-of-the-art methods and at the same time enjoys good interpretability.
Cite
Text
Zhao et al. "A Document-Grounded Matching Network for Response Selection in Retrieval-Based Chatbots." International Joint Conference on Artificial Intelligence, 2019. doi:10.24963/IJCAI.2019/756Markdown
[Zhao et al. "A Document-Grounded Matching Network for Response Selection in Retrieval-Based Chatbots." International Joint Conference on Artificial Intelligence, 2019.](https://mlanthology.org/ijcai/2019/zhao2019ijcai-document/) doi:10.24963/IJCAI.2019/756BibTeX
@inproceedings{zhao2019ijcai-document,
title = {{A Document-Grounded Matching Network for Response Selection in Retrieval-Based Chatbots}},
author = {Zhao, Xueliang and Tao, Chongyang and Wu, Wei and Xu, Can and Zhao, Dongyan and Yan, Rui},
booktitle = {International Joint Conference on Artificial Intelligence},
year = {2019},
pages = {5443-5449},
doi = {10.24963/IJCAI.2019/756},
url = {https://mlanthology.org/ijcai/2019/zhao2019ijcai-document/}
}