A knowledge based system for mining association rules for video categories

Video data mining is a challenging research area due to interesting nature of unstructured video data. Generating association rules between items in a large video database plays a significant role in the video mining research areas. Applications of video association mining are not limited to the domains of surveillance, meetings, news broadcast, sports, video on demand (VOD), telemedicine, biomedical engineering and as well as online media collections. This paper concentrates on a knowledge-based system to generate association rules for selecting video categories using “Belief Rule Base (BRB)”. It has been shown that the system is efficient than traditional association rule mining.