Reference no: EM133320160

The inference of association rules when all terms are "positive / affirmative", that is, an item is present. The case of having negations was not discussed. Develop approaches that can simultaneously use affirmative and negative terms (such as a hypothetical itemset of the form {A, B, not(D), H, not(W)}). That is, now terms such as "not(D)", and "not(W)", indicate the absence of items D and W, respectively. Respond in terms of the following three tasks:

Task 1: Develop a revised a-priory method that can handle such problems.

Task 2: Describe the appropriate monotonicity properties when identifying frequent itemset and association rules from a given frequent itemset for the new problem.

Task 3: Compare the methods when only positive terms are used versus when positive and negated terms are used. Use small examples to demonstrate your methods. You define the appropriate threshold values. Describe all your algorithms in terms of pseudo-code too.