Inducing Semantic Representation from Text by Jointly Predicting and Factorizing Relations

TL;DR

This paper introduces a joint model that combines semantic role labeling with tensor factorization to induce semantic representations from text without relying on prior linguistic knowledge.

Contribution

It presents a novel joint approach that integrates role prediction and relation factorization, achieving competitive results without using prior linguistic resources.

Findings

Achieves performance comparable to state-of-the-art role induction methods

Induced roles align closely with annotated semantic roles

Operates effectively without prior linguistic knowledge

Abstract

In this work, we propose a new method to integrate two recent lines of work: unsupervised induction of shallow semantics (e.g., semantic roles) and factorization of relations in text and knowledge bases. 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, even though, unlike these previous approaches, we do not incorporate any prior linguistic knowledge about the language.