Zero-Shot Relation Extraction via Reading Comprehension
This approach addresses the problem of extracting relations from text without labeled examples for new types, offering a novel method for slot-filling tasks.
The paper tackles relation extraction by reducing it to answering reading comprehension questions, enabling zero-shot learning for unseen relation types with lower accuracy and high accuracy for known types.
We show that relation extraction can be reduced to answering simple reading comprehension questions, by associating one or more natural-language questions with each relation slot. This reduction has several advantages: we can (1) learn relation-extraction models by extending recent neural reading-comprehension techniques, (2) build very large training sets for those models by combining relation-specific crowd-sourced questions with distant supervision, and even (3) do zero-shot learning by extracting new relation types that are only specified at test-time, for which we have no labeled training examples. Experiments on a Wikipedia slot-filling task demonstrate that the approach can generalize to new questions for known relation types with high accuracy, and that zero-shot generalization to unseen relation types is possible, at lower accuracy levels, setting the bar for future work on this task.