Automatic web translators as part of a multilingual question-answering (QA) system: translation of questions

Abstract

The traditional model of information retrieval entails some implicit restrictions, including: a) the assumption that users search for documents, not answers; and that the documents per se will respond to and satisfy the query, and b) the assumption that the queries and the document that will satisfy the particular informational need are written in the same language. However, many times users will need specific data in response to the queries put forth. Cross-language question-answering systems (QA) can be the solution, as they pursue the search for a minimal fragment of text—not a complete document—that applies to the query, regardless of the language in which the question is formulated or the language in which the answer is found. Cross-language QA calls for some sort of underlying translating process. At present there are many types of software for natural language translation, several of them available online for free. In this paper we describe the main features of the multilingual Question-Answering (QA) systems, and then analyze the effectiveness of the translations obtained through three of the most popular online translating tools (Google Translator, Promt and Worldlingo). The methodology used for evaluation, on the basis of automatic and subjective measures, is specifically oriented here to obtain a translation that will serve as input in a QA system. The results obtained 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.