Multi-Rule Based Ensemble Feature Selection Model for Sarcasm Type Detection in Twitter

Algorithm 5

Fuzzy algorithm for sarcastic type detection.

	Input: Feature Set
	Output: Type of Sarcasm
	(1) For every fuzzy input variable fi in selected feature set, F
	a. Define fuzzy set corresponding to the crisp values
	b. Define the membership functions
	(2) Define fuzzy set for the consequent i.e. Type of Sarcasm
	(3) Define membership function for the output variable
	(4) For in type of sarcasm
	a. Define fuzzy rules
	(5) Perform a control system simulation using the defined rules
	(6) For tweet ti ∈ list of tweets T
	a. Compute the fuzzy value of type of sarcasm
	b. Compute the defuzzified value using centroid method
	(7) Defuzzified value will determine the type of sarcasm to which each text belongs to.