Toward a neural aggregated search model for semi-structured documents

Abstract

In this paper, we are interested in content-oriented XML information retrieval. Our goal is to revisit the granularity of the unit to be returned. More precisely, instead of returning the whole document or a list of disjoint elements of a document, as it is usually done in the most XML information retrieval systems, we attempt to build the best elements aggregation (set of non-redundant elements) which is likely to be relevant to a query composed of key words. Our approach is based on Kohonen self-organizing maps. Kohonen self-organizing map allows an automatic classification of XML elements producing density map that form the foundations of aggregated search.