CokeBERT: Contextual Knowledge Selection and Embedding towards Enhanced Pre-Trained Language Models
This work addresses the issue of redundant and ambiguous knowledge in knowledge-enhanced language models for NLP researchers and practitioners, representing an incremental advancement over existing methods.
The paper tackles the problem of static knowledge integration in pre-trained language models by proposing Coke, a framework that dynamically selects and embeds contextual knowledge from knowledge graphs based on the input text, achieving performance improvements on knowledge-driven NLP tasks.
Several recent efforts have been devoted to enhancing pre-trained language models (PLMs) by utilizing extra heterogeneous knowledge in knowledge graphs (KGs) and achieved consistent improvements on various knowledge-driven NLP tasks. However, most of these knowledge-enhanced PLMs embed static sub-graphs of KGs ("knowledge context"), regardless of that the knowledge required by PLMs may change dynamically according to specific text ("textual context"). In this paper, we propose a novel framework named Coke to dynamically select contextual knowledge and embed knowledge context according to textual context for PLMs, which can avoid the effect of redundant and ambiguous knowledge in KGs that cannot match the input text. Our experimental results show that Coke outperforms various baselines on typical knowledge-driven NLP tasks, indicating the effectiveness of utilizing dynamic knowledge context for language understanding. Besides the performance improvements, the dynamically selected knowledge in Coke can describe the semantics of text-related knowledge in a more interpretable form than the conventional PLMs. Our source code and datasets will be available to provide more details for Coke.