Work

Multi-Cell Compositional LSTM for NER Domain Adaptation

Year: 2020

Type: conference-paper

Abstract: Cross-domain NER is a challenging yet practical problem.Entity mentions can be highly different across domains.However, the correlations between entity types can be relatively more stable across domains.We investigate a multi-cell compositional LSTM structure for multi-task learning, modeling each entity type using a separate cell state.With the help of entity typed units, cross-domain knowledge transfer can be made in an entity type level.Theoretically, the resulting distinct feature distributions for each entity type make it more powerful for cross-domain transfer.Empirically, experiments on four few-shot and zeroshot datasets show our method significantly outperforms a series of multi-task learning methods and achieves the best results.

Source: Annual Meeting of the Association for Computational Linguistics (ACL)

DOI: 10.18653/v1/2020.acl-main.524

Authors: Chen Jia, Yue Zhang

Institutions: Fudan University, Westlake University

Language: English

Topic: Topic Modeling

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: hybrid

Funders: Westlake University, National Natural Science Foundation of China

Award: G8869030430