DiVAS: Digital Video Audio Sketch a Prototype for Cross-Media Capture, Sharing, and Reuse of Rich Contextual Knowledge

People generate and develop concepts in informal settings through gesture language, verbal discourse, and sketching. Current knowledge capture and reuse solutions are unable to utilize the relevance embedded in these multiple multimodal streams of communication. As a result, most of the archived content becomes hard to understand and reuse. This paper presents DiVAS (Digital Video-Audio-Sketch) a system for cross-media knowledge capture, search, retrieval and re-use of rich, unstructured, multimedia digital content stored in large corporate repositories. This is achieved through seamless transformation of analog activities such as gestures, discourse, and sketching that take place during creative and informal communicative events among stakeholders engaged in building projects, into integrated digital video-audio-sketch episodes. DiVAS presents a cross-media semantic analysis and data mining methodology of indexed digital video-audio-sketch footage that captures the creative human activities of concept generation and problem solving. Together, they provide a macro-micro index to large enterprise archives of rich, multimedia, and unstructured content. Knowledge re-use is facilitated through contextual exploration and understanding of the rich content that is retrieved.