Probabilistic Coreference in Information Extraction

TL;DR

This paper explores methods to assign probabilistic distributions to coreference relationships in information extraction, enabling better integration of conflicting data from multiple sources.

Contribution

It introduces approaches for estimating probability distributions over coreference sets within an information extraction framework, addressing a key need for probabilistic outputs.

Findings

Initial experiments demonstrate feasibility of probabilistic coreference assignment

Approaches improve integration of conflicting information

Framework supports downstream fusion of extracted data

Abstract

Certain applications require that the output of an information extraction system be probabilistic, so that a downstream system can reliably fuse the output with possibly contradictory information from other sources. In this paper we consider the problem of assigning a probability distribution to alternative sets of coreference relationships among entity descriptions. We present the results of initial experiments with several approaches to estimating such distributions in an application using SRI's FASTUS information extraction system.