Imposing Label-Relational Inductive Bias for Extremely Fine-Grained Entity Typing
This addresses the challenge of fine-grained entity typing for NLP applications where type sets are not restricted by knowledge base schemas, representing an incremental advance over existing methods.
The paper tackled the problem of entity typing in open scenarios with free-form types lacking a predefined hierarchy, achieving a 15.3% relative F1 improvement and higher recall while maintaining precision on a dataset with over 10,000 types.
Existing entity typing systems usually exploit the type hierarchy provided by knowledge base (KB) schema to model label correlations and thus improve the overall performance. Such techniques, however, are not directly applicable to more open and practical scenarios where the type set is not restricted by KB schema and includes a vast number of free-form types. To model the underly-ing label correlations without access to manually annotated label structures, we introduce a novel label-relational inductive bias, represented by a graph propagation layer that effectively encodes both global label co-occurrence statistics and word-level similarities.On a large dataset with over 10,000 free-form types, the graph-enhanced model equipped with an attention-based matching module is able to achieve a much higher recall score while maintaining a high-level precision. Specifically, it achieves a 15.3% relative F1 improvement and also less inconsistency in the outputs. We further show that a simple modification of our proposed graph layer can also improve the performance on a conventional and widely-tested dataset that only includes KB-schema types.