Molecular sequence alignment for extracting answers for where-typed questions from google snippets

Abstract

This work presents a strategy to answer questions about locations based on a character alignment algorithm that ranks Google snippets according to their similarity with the query. Answer candidates are extracted by using a Name-Entity Recognizer and then ranked according to their position in this new rank. Results concerning questions on location of monuments and cities around the world are discussed. © Springer-Verlag Berlin Heidelberg 2006.