A Probabilistic Model with Commonsense Constraints for Pattern-based Temporal Fact Extraction
This work addresses the challenge of unreliable pattern-based extraction for temporal facts in news data, offering an incremental improvement with novel modeling of time signals and commonsense constraints.
The authors tackled the problem of extracting accurate temporal facts from news data using textual patterns by proposing a probabilistic graphical model that infers true facts and pattern reliability without supervision, incorporating temporal signals and commonsense constraints, and demonstrated significant performance improvements over existing methods.
Textual patterns (e.g., Country's president Person) are specified and/or generated for extracting factual information from unstructured data. Pattern-based information extraction methods have been recognized for their efficiency and transferability. However, not every pattern is reliable: A major challenge is to derive the most complete and accurate facts from diverse and sometimes conflicting extractions. In this work, we propose a probabilistic graphical model which formulates fact extraction in a generative process. It automatically infers true facts and pattern reliability without any supervision. It has two novel designs specially for temporal facts: (1) it models pattern reliability on two types of time signals, including temporal tag in text and text generation time; (2) it models commonsense constraints as observable variables. Experimental results demonstrate that our model significantly outperforms existing methods on extracting true temporal facts from news data.