Injecting Domain Knowledge in Language Models for Task-Oriented Dialogue Systems
This addresses the need for better integration of small, modifiable knowledge bases in real-world dialogue systems, though it is incremental as it builds on existing adapter methods.
The paper tackles the problem of pre-trained language models lacking domain-specific knowledge for task-oriented dialogue systems by injecting such knowledge using light-weight adapters, resulting in improvements over strong baselines on knowledge probing and response generation tasks.
Pre-trained language models (PLM) have advanced the state-of-the-art across NLP applications, but lack domain-specific knowledge that does not naturally occur in pre-training data. Previous studies augmented PLMs with symbolic knowledge for different downstream NLP tasks. However, knowledge bases (KBs) utilized in these studies are usually large-scale and static, in contrast to small, domain-specific, and modifiable knowledge bases that are prominent in real-world task-oriented dialogue (TOD) systems. In this paper, we showcase the advantages of injecting domain-specific knowledge prior to fine-tuning on TOD tasks. To this end, we utilize light-weight adapters that can be easily integrated with PLMs and serve as a repository for facts learned from different KBs. To measure the efficacy of proposed knowledge injection methods, we introduce Knowledge Probing using Response Selection (KPRS) -- a probe designed specifically for TOD models. Experiments on KPRS and the response generation task show improvements of knowledge injection with adapters over strong baselines.