The ontology-cored emotional semantic search model

Abstract

An ontology-cored emotional semantic retrieval model was proposed and constructed in order to overcome the current situation that it is difficult to implement emotional semantic search in the course of image searching. A method Combines Mpeg-7, theory of concept lattices and ontology construction together to construct the kernel ontology library was put forward. The difficulty of this method lies on how to integrate the Mpeg-7 standard descriptor with image emotional ontology properties, and how to auto-generate new concept results. Related experiments were carried out of which the results validated the feasibility of this model on image emotional semantic search. © 2010 Springer-Verlag Berlin Heidelberg.