paper-with-me

홈 › Papers

Contextual Fine-to-Coarse Distillation for Coarse-grained Response Selection in Open-Domain Conversations

2021-09-24 · ACL 2022 5 · Wei Chen, Yeyun Gong, Can Xu, Huang Hu, Bolun Yao, Zhongyu Wei, Zhihao Fan, Xiaowu Hu, Bartuer Zhou, Biao Cheng, Daxin Jiang, Nan Duan

We study the problem of coarse-grained response selection in retrieval-based dialogue systems. The problem is equally important with fine-grained response selection, but is less explored in existing literature. In this paper, we propose a Contextual Fine-to-Coarse (CFC) distilled model for coarse-grained response selection in open-domain conversations. In our CFC model, dense representations of query, candidate response and corresponding context is learned based on the multi-tower architecture, and more expressive knowledge learned from the one-tower architecture (fine-grained) is distilled into the multi-tower architecture (coarse-grained) to enhance the performance of the retriever. To evaluate the performance of our proposed model, we construct two new datasets based on the Reddit comments dump and Twitter corpus. Extensive experimental results on the two datasets show that the proposed methods achieve a significant improvement over all evaluation metrics compared with traditional baseline methods.

📄 PDF Abstract BibTeX arXiv:2109.13087

Code (1)

lemuria-wchen/CFC 공식 구현

Tasks

Retrieval

Similar Papers 제목 키워드 기반

Contextual Fine-to-Coarse Distillation for Coarse-grained Response Selection in Open-Domain Conversations

2021-11-16 · ACL ARR November 2021 11 · Anonymous

We study the problem of coarse-grained response selection in retrieval-based dialogue systems. The problem is equally important with fine-grained response selection, but is less explored in existing literature. In this p…

Retrieval

Improving Span-based Question Answering Systems with Coarsely Labeled Data

2018-11-05 · Hao Cheng, Ming-Wei Chang, Kenton Lee, Ankur Parikh 외

We study approaches to improve fine-grained short answer Question Answering models by integrating coarse-grained data annotated for paragraph-level relevance and show that coarsely annotated data can bring significant pe…

Multi-Task LearningQuestion Answering

Coarse-to-Fine Domain Incremental Learning with Attentive Distillation for Mining Footprint Segmentation in Multispectral Imagery

2026-05-23 · Alif Tri Handoyo, Vincent C. S. Lee, Rizka Widyarini Purwanto, Alex M. Lechner 외 arxiv

Automatically mapping and segmenting global mining footprints using remote sensing and deep learning is critical for monitoring the socio-environmental risks and impacts of mining, yet its progress is hindered by the sca…

Incremental LearningDomain Adaptation

Progressive refinement: a method of coarse-to-fine image parsing using stacked network

2018-04-23 · Jiagao Hu, Zhengxing Sun, Yunhan Sun, Jinlong Shi

To parse images into fine-grained semantic parts, the complex fine-grained elements will put it in trouble when using off-the-shelf semantic segmentation networks. In this paper, for image parsing task, we propose to par…

Face ParsingHuman ParsingSegmentationSemantic Segmentation

AMBERT: A Pre-trained Language Model with Multi-Grained Tokenization

2020-08-27 · Findings (ACL) 2021 8 · Xinsong Zhang, Pengshuai Li, Hang Li

Pre-trained language models such as BERT have exhibited remarkable performances in many tasks in natural language understanding (NLU). The tokens in the models are usually fine-grained in the sense that for languages lik…

Language ModelingLanguage ModellingNatural Language Understanding