Application of clustering algorithm on TV programmes preference grouping of subscribers

With the development of digital cable interactive business and the diversification of the customers´ demand, grouping TV programmes based on preferences of users effectively is vital for market segmentation and differentiation. The study summarizes the main principle and characteristic of clustering algorithm, and uses K-Means algorithm to show TV programmes preference grouping based on 52392 subscribers in a given area. Overall, the results show that K-Means algorithm is a better method to mine the data of television audience behavior; the clustering result could be a great guidance and the study lays a good foundation for analyzing TV user behavior.