Generation of constraint based sequential rules with Trulegrowth Algorithm

Sequential rules are useful for prediction of an event with a probability called as confidence. It uses Association rule mining principle for the discovery of sequential rules. Various algorithms have designed for mining sequential rules efficiently. The generation of a large number of rules becomes inefficient for analysis and may lead to the wrong prediction. The constraint based sequential rule generation system generates sequential rules with use of the TRuleGrowth algorithm with an addition of user interested constraints like length, attribute, regular expression, gap, duration etc. which can be used for better analysis and for fine prediction and decision making. Experimental study shows that no of rules generated, time and memory required to generate the rules is reduced with the constraint based sequential rule generation system.