NLEL at RespubliQA 2010 Santiago Correa, Davide Buscaldi, and Paolo Rosso Natural Language Engineering Lab., ELiRF Universidad Polit´ ecnica de Valencia, Camino de Vera s/n, Valencia, Espa˜ na santcg@gmail.com{scorrea,dbuscaldi,prosso}@dsic.upv.es {dbuscaldi,prosso}@dsic.upv.es http://www.dsic.upv.es/grupos/nle Abstract. This report describes the participation of the NLEL Lab. from the Universidad Polit´ ecnica of Valencia to the RespubliQA task at CLEF 2010. The system designed for this participation is based on the one used in our previous participation, with some modifications required in order to adapt it to the new guidelines. The system participated to both the “Paragraph Selection” (PS) and “Answer Selection” (AS) sub- tasks. Keywords: Question Answering, n-gram based Passage Retrieval 1 Introduction The participation to the PS sub-task was centered around the JIRS n-gram based passage retrieval system [6]. In order to participate in the AS sub-task, it was necessary to integrate into the system an Answer Extraction module, which was developed originally for the QUASAR QA system [5], which participated in past CLEF-QA editions, from 2005 to 2007. In the following sections we describe the characteristics of the QA system in both PS and AS configurations. 2 JIRS Passage Retrieval System JIRS 1 is an n-gram based passage retrieval system that has been developed specifically for the Question Answering task. An n-gram is a sequence of n adjacent terms extracted from a sentence or a question. JIRS is based on the premise that in a sufficiently large document collection, question n-grams should appear near the answer at least once. JIRS represents the core of the system, since it was used both in the PS and AS sub-tasks. The architecture of JIRS is shown in Figure 1. The user question is passed to a search engine that returns relevant snippets of a documents collection in which relevant terms from the question occur. The n -gram extraction module will return all the n -grams of size 1 to n, where n is the number of terms of the question. This process is done both for the question and for each of the snippets 1 http://sourceforge.net/projects/jirs/