Analysis of automatic translation of questions for question-answering systems

Abstract

Introduction. Multilingual question-answering systems can provide users with specific data in response to queries by searching for a minimal fragment of text that applies to the query, regardless of the language in which the question is formulated and the answer is found. The aim of this paper is to analyse the automatic translation of questions (intended as queries input to a cross-language, question-answering system) from German and French into the Spanish language. Method. The methodology used for evaluation, based on automatic and subjective measures, appraises whether the translation will serve as input to a system. That is, does the question retain its validity and fulfil its function, allowing a proper response to be found? Analysis. The main features of multilingual question-answering systems are described and then we analyse the effectiveness of the translations achieved through three popular online translating tools: Google Translator, Promt and Worldlingo. Results. Our findings serve to identify which is the most reliable translator for both pairs of languages overall. However, an even more reliable option would be to use two different translators, depending on which of the two source languages is being dealt with. Conclusions. The results contribute to the realm of innovative search systems by enhancing our understanding of online translators and their potential in the context of multilingual information retrieval.