Unsupervised Induction of Semantic Roles within a Reconstruction-Error Minimization Framework

TL;DR

This paper presents an unsupervised method for semantic role labeling that jointly learns role prediction and argument reconstruction, achieving competitive results without prior linguistic knowledge.

Contribution

It introduces a novel joint model combining role prediction and argument reconstruction, demonstrating effective unsupervised semantic role induction across languages.

Findings

Achieves performance comparable to state-of-the-art supervised methods

Does not require prior linguistic knowledge

Works effectively for English and German

Abstract

We introduce a new approach to unsupervised estimation of feature-rich semantic role labeling models. Our model consists of two components: (1) an encoding component: a semantic role labeling model which predicts roles given a rich set of syntactic and lexical features; (2) a reconstruction component: a tensor factorization model which relies on roles to predict argument fillers. When the components are estimated jointly to minimize errors in argument reconstruction, the induced roles largely correspond to roles defined in annotated resources. Our method performs on par with most accurate role induction methods on English and German, even though, unlike these previous approaches, we do not incorporate any prior linguistic knowledge about the languages.