Web image annotation using word co-occurrence and association analysis

Abstract

With the popularization of Web and image devices, there are more and more digital images available on the Internet. How to effectively organize and manage these web images becomes a critical issue. In this paper, a novel approach that employs the relationships between words to achieve web image annotations is proposed. First, the related textual information associated with web images is identified as the candidate annotations for web images. Second, the word co-occurrence is utilized to eliminate irrelevant keywords for improving the annotation accuracy. Then, the keyword-based association analysis is exploited to further discover possible semantics for enriching the final annotations of web images. Finally, experimental results demonstrate the effectiveness of the proposed method