Training and Scaling Preference Functions for Disambiguation

TL;DR

This paper introduces an automatic method for optimizing preference function weights in disambiguation tasks, demonstrating improved performance over hand-tuned factors, especially for functions based on semantic lexical collocations.

Contribution

It proposes a novel automatic scaling approach for preference functions in disambiguation, with a focus on semantic lexical collocations, outperforming traditional methods.

Findings

Automatic scaling improves disambiguation accuracy.

Semantic lexical collocation functions vary in effectiveness.

A new function outperforms mutual information-based functions.

Abstract

We present an automatic method for weighting the contributions of preference functions used in disambiguation. Initial scaling factors are derived as the solution to a least-squares minimization problem, and improvements are then made by hill-climbing. The method is applied to disambiguating sentences in the ATIS (Air Travel Information System) corpus, and the performance of the resulting scaling factors is compared with hand-tuned factors. We then focus on one class of preference function, those based on semantic lexical collocations. Experimental results are presented showing that such functions vary considerably in selecting correct analyses. In particular we define a function that performs significantly better than ones based on mutual information and likelihood ratios of lexical associations.