• Data includes images, video, audio, graphs, etc.
• Mostly embedded into Web pages, often via hyperlinks
• Increasing demand for effective methods to organise and retrieve multimedia data
• Web page layout mining can be used to find multimedia blocks
• Example: Classify images
• Use VIPS to identify multimedia blocks in a Web page
• Use textual description around images for classification/categorisation
• Use block-level link analysis (rather than page level link analysis such as Google’s PageRank)