As mentioned we can get multiple length association rules but support and confidence threshold values is main deciding factor to be considered. apriori algorithm is vastly used and with combination of partitioning, hashing, sampling ... for the optimization purpose.