Nowadays the popularity of tag clouds in websites is increased notably, but its generation is criticized because its lack of control causes it to be more likely to produce inconsistent and redundant results. It is well known that if tags are freely chosen (instead of taken from a given set of terms), synonyms (multiple tags for the same meaning), normalization of words and even, heterogeneity of users are likely to arise, lowering the efficiency of content indexing and searching contents. To solve this problem, we have designed the Maximum Similarity Measure (MaSiMe) a dynamic and flexible similarity measure that is able to take into account and optimize several considerations of the user who wishes to obtain a free-of-redundancies tag cloud. Moreover, we include an algorithm to effectively compute the measure and a parametric study to determine the best configuration for this algorithm.
