GraphPrompt: Biomedical Entity Normalization Using Graph-based Prompt Templates

Abstract

Biomedical entity normalization unifies the language across biomedical experiments and studies, and further enables us to obtain a holistic view of life sciences. Current approaches mainly study the normalization of more standardized entities such as diseases and drugs, while disregarding the more ambiguous but crucial entities such as pathways, functions and cell types, hindering their real-world applications. To achieve biomedical entity normalization on these under-explored entities, we first introduce an expert-curated dataset OBO-syn encompassing 70 different types of entities and 2 million curated entity-synonym pairs. To utilize the unique graph structure in this dataset, we propose GraphPrompt, a promptbased learning approach that creates prompt templates according to the graphs. Graph-Prompt obtained 41.0% and 29.9% improvement on zero-shot and few-shot settings respectively, indicating the effectiveness of these graph-based prompt templates. We envision that our method GraphPrompt and OBO-syn dataset can be broadly applied to graph-based NLP tasks, and serve as the basis for analyzing diverse and accumulating biomedical data.

Graph Structure Prompt Learning: A Novel Methodology to Improve Performance of Graph Neural Networks

“…Prompt-Based Learning. In recent two years, prompt-based learning methods [42,43,45] have shown strong capabilities to improve the downstream tasks of NLP. Schick et al [42] utilize prompt templates to provide label hints of models for classification and generation language tasks.…”

“…In this way, structural feature learning of graphs can be enhanced and may benefit downstream tasks. Recently, prompt-based tuning methods [42,43,44,45] in natural language processing (NLP) represent a powerful performance by bridging gaps between the pre-training stage and downstream task fine-tuning stages. For instance, given an input sentence of "The story of the movie is well arranged.…”