The application of multimodal data mining:A survey

Nowadays, multimodal data available to us through the Internet and other electronic media are explosively increasing both in number and in variety. Efficiently analyzing these data will save people a large amout of time. However, due to the property of multimodal data, methods used to uni-modal data usually perform ineffectively here. To handle such massive multimodal data for various purposes, new technologies are in need of development. Data Mining helps people find a potentially useful knowledge from Massive Data, and play an important role in many fields. Data mining could only handle simple text data in the past. After several decades of development, data mining can deal with the multimedia data with complex structure now. Multimedia data is a sort of multimodal data. Some fusion and cross-media data mining algorithms can be used to mine multimodal data effectively. Based on the data mining approach to the task of analyzing multimodal data, this paper briefly introduces several methods fitting for multimodal data mining.