Generating temporal association rules for infrequent items in concept level video datasets

K.Kumar and P.Sudhakar

Due to the increasing rate of video data over the World Wide Web and smart phones, it is becoming very essential to extract useful information from visual data. The video data consists lot of objects which in turn contain important visual information. Video data also contains an amount of direct as well as indirect or hidden information about its objects. The users can access direct information from video by viewing it. To access hidden information from the video mining techniques such as classifications, clustering, regression, outlier detection and association rules etc, can be applied. We focus on concept level video associations and propose an algorithm to generate temporal association rule for the same.

Download PDF: 
Select Volume: