Abstract: In era of big data, we need more efficient algorithm and application of data mining. As a main direction, the algorithm of association rules can discover frequent item sets and association knowledge in a large amount of data. Apriori algorithm is a classical algorithm of association rules, based on this research, we find out the weakness of it and get improved methods for massive data. On its application in the customer viewing data, the improved algorithm increases the efficiency of data mining. Also after getting the mining results, we propose new topics for post application.

Key words: association rules, Apriori algorithm, viewing behavior, data processing