An Information Retrieval Model for the Semantic Web

Exploring the metadata associated with documents in the Semantic Web is a way to increase the precision of information retrieval systems. Systems have been established so far failed to overcome fully the limitations of search based on keywords. Such systems are built from variations of classic models that represent information by keywords and work upon statistical correlations. This work proposes an information retrieval model to find information items with similar semantic content that a given user’s query. The information items internal representation is based on user interest groups, called "semantic cases". The model also defines a similarity measure for ordering the results based on semantic distance between semantic cases items.

1 Introduction

Several information retrieval process models, such the Boolean, the vector space and the probabilistic model have been proposed to cover the activities and technical user queries as well as storage and retrieval of information items from unstructured sources [3]. Classic models represent the documents with a set of keywords extracted from text and propose different approaches to retrieval and presentation of retrieved information items sorted according to their relevance.