Learning Scripts as Hidden Markov Models

TL;DR

This paper introduces a formal framework for narrative scripts using Hidden Markov Models, enabling robust inference and learning, and demonstrates its effectiveness in predicting missing events in natural datasets.

Contribution

It is the first to formalize scripts as Hidden Markov Models, providing algorithms for structure and parameter learning with superior performance over baselines.

Findings

The HMM-based framework outperforms baseline models in event prediction.

The developed algorithms effectively learn script structures from natural data.

Robust inference algorithms improve narrative understanding tasks.

Abstract

Scripts have been proposed to model the stereotypical event sequences found in narratives. They can be applied to make a variety of inferences including filling gaps in the narratives and resolving ambiguous references. This paper proposes the first formal framework for scripts based on Hidden Markov Models (HMMs). Our framework supports robust inference and learning algorithms, which are lacking in previous clustering models. We develop an algorithm for structure and parameter learning based on Expectation Maximization and evaluate it on a number of natural datasets. The results show that our algorithm is superior to several informed baselines for predicting missing events in partial observation sequences.