Back Deduction Based Testing for Word Sense Disambiguation Ability of Machine Translation Systems

Abstract

Machine translation systems have penetrated our daily lives, providing translation services from source language to target language to millions of users online daily. Word Sense Disambiguation (WSD) is one of the essential functional requirements of machine translation systems, which aims to determine the exact sense of polysemes in the given context. Commercial machine translation systems (e.g., Google Translate) have been shown to fail in identifying the proper sense and consequently cause translation errors. However, to our knowledge, no prior studies focus on testing such WSD bugs for machine translation systems. To tackle this challenge, we propose a novel testing method Back Deduction based Testing for Word Sense Disambiguation (BDTD). Our method's main idea is to obtain the hidden senses of source words via back deduction from the target language, i.e., employ translation words in the target language to deduce senses of original words identified in the translation procedure. To evaluate BDTD, we conduct an extensive empirical study with millions of sentences under three popular translators, including Google Translate and Bing Microsoft Translator. The experimental results indicate that BDTD can identify a considerable number of WSD bugs with high accuracy, more than 80%, under all three translators.