Word semantics for information retrieval: moving one step closer to the Semantic Web

The goal of the Semantic Web is to create a new form of Web content meaningful to computers. The Semantic Web aims to provide greater functionality, via intelligent tools such as information extractors, brokers, reasoning services or question answering systems. Semantics can be addressed at several levels. In this paper, we focus on the lowest level-word semantics on which other higher levels such as concept, paragraph, or document levels can be based upon. This model, which we call Word Semantics (WS), does not include the rich set of tags proposed by the XML/RDF standards. Nevertheless, this simpler WS format comes with a big advantage: it is possible with existing technologies and resources. Practically, this new model relies on understanding word meanings, identifying important named entities such as person, organization and others, and linking all this information via an external general purpose ontology, namely WordNet. With these features, we regard the WS model as a short but strong step toward the long term goal of a Semantic Web.