Modified Lesk Algorithm

function LESK (word, sentence) returns best sense of word
	best sense <- most frequent use for word
	max_overlap <- 0
	context <- set of words in sentence
	intersection, synset_no, meaning = EXTRAOVERLAP (word, sentence)
	for each sense in sense of word do
		signature <- set of words in the gloss and examples of sense
		overlap <- COMPUTEOVERLAP (signature, context)
		if overlap > max_overlap then
			max_overlap <- overlap
	if  intersection > max_overlap and intersection > 0 do
		best_sense <- word.synset_no
end return (best_sense)