Personalized indexing for heterogeneous multimedia data

The work presented in this article fall under the scope of multimedia documents coming from a heterogeneous corpus that its aims to improve the quality of semantic search of information. We propose a unified indexing model based on the integration of user profile to the indexing process. The general idea of our proposal is to operate the common concepts between the representation of a document and the definition of a user via his profile, these two elements will be added as additional indexing entities to enrich the heterogeneous corpus database indexes. We believe that these new concepts will enable the improvement of the automatic judging of documents relating to user profile. We have developed IRONTO domain ontology allowing annotation of documents in the field of information retrieval. We will present also the tool developed validating the proposed model. An evaluation phase aims at comparing the execution result according to two indexing modes: one without integration of the user profile and the other with the integration of the user profile using our unified model. The results show that the use of user profile in the process of semantic indexing improves significantly the performance of research in relation to the first indexing mode.