Análisis documental de noticias de prensa en sistemas de información factual

Abstract

Digital news libraries provide users with great amounts of information both in full text and references, that usually require a great amount of additional time for searching for factual data: subjects, consequences, dates, places, causes, backgrounds...The current research proposes a pragmatic procedure of analysis and retrieval focusing on press news items and based on pronominal request modes formulated by end-users. This has a clear potential for extrapolation to other journalistic genres, such as interviews or news reports, which would have to be separately and specifically addressed. This study begins with the construction of macro-propositions based on a corpus of news items and following rules that are both practical and easy to reproduce. The goal is to obtain a means for accessing press libraries and archives that is both standardized and automatic. © 2014 CSIC.