Concept-based visual information management with large lexical corpus

Most users want to find visual information based on the semantics of visual contents such as a name of person, semantic relations, an action happening in a scene,…etc. However, techniques for content-based image or video retrieval are not mature enough to recognize visual semantic completely, whereas retrieval based on color, size, texture and shape are within the state of the art. Therefore, smart ways to manage textual annotations in visual information retrieval are necessary. In this paper, a framework for integration of textual and visual content searching mechanism is presented. The proposed framework includes ontology-based semantic query processing through efficient semantic similarity measurement. A new conceptual similarity distance measure between two conceptual entities in a large taxonomy structure is proposed and its efficiency is demonstrated. With the proposed method, an information retrieval system can benefit such as (1) reduction of the number of trial-and-errors to find correct keywords, (2) Improvement of precision rates by eliminating the semantic heterogeneity in description, and (3) Improvement of recall rates through precise modeling of concepts and their relations.