The Capture, Storage, Sharing, Organizing, Retrieval, and Use of knowledge dominate most socio-economic activities in our society. Most of the knowledge in the world is initially captured and stays in the form of experiences in different sensing modalities. Current technology can address knowledge in text because in text the experiential data is converted to symbols by humans. Converting sensory data to symbols in computer systems has been difficult, primarily due to our inability to formally represent and effectively model the appropriate context within which the multimedia sensory data should be interpreted. This problem of integrating continuous multimedia and symbolic data becomes more urgent as multimedia data is becoming common, and the resulting social applications on the Web are required to deal with semantics of multimedia data. Clearly, the collection of searchable multimedia experiences will facilitate progress in sciences and the quality of human life in every part of the world, across all economies and cultures. In this paper, we will present challenges offered by semantics of multimedia data, review emerging semantic web approaches towards addressing these challenges, and present some example applications that are being developed to address these emerging challenges.