Title :
An information theoretic similarity measure for unified multimedia document retrieval
Author :
Pushpalatha, K. ; Ananthanarayana, V.S.
Author_Institution :
Dept. of Inf. Technol., Nat. Inst. of Technol. Karnataka, Surathkal, India
Abstract :
Due to the rapid evolution in multimedia technology, the multimedia data have been growing at a phenomenal rate. With the enormous amount of multimedia data, the richness of its information has raised the demand for sophisticated multimedia knowledge discovery systems. Multimedia documents requires distinct type of processing and knowledge discovery methods, due to the distinct characteristics of multimedia data. However, the individual processing and analyzing each type of multimedia data may make the system more tedious and complex. Also, the correlations between the contents of multimedia document are not considered in the knowledge extraction process. When the multimodal objects are represented in the unimodal domain, similar processing and knowledge discovery methods can be used. In this paper, a domain converting method is proposed with the aim to represent the multimedia document of multimodal objects as multimedia document of signal objects. In order to find the similarity between the multimedia documents, an information theory based similarity measure is proposed. Evaluation of the proposed framework has been performed by experimenting the system for the retrieval of multimedia documents. The proposed domain converting method presents the multimodal document in the unified representation. Experimental results of multimedia document retrieval shows better document retrieval rate with the proposed similarity measure.
Keywords :
data mining; information retrieval; information theory; multimedia computing; domain converting method; information theoretic similarity measure; knowledge extraction process; multimedia data; multimedia knowledge discovery systems; multimedia technology; multimodal objects; signal objects; unified multimedia document retrieval; Correlation; Feature extraction; Information theory; Media; Multimedia communication; Vectors; Visualization; Information Retrieval; Information Theory; Multimedia; Signal Object; Similarity Measure;
Conference_Titel :
Information and Automation for Sustainability (ICIAfS), 2014 7th International Conference on
DOI :
10.1109/ICIAFS.2014.7069625