Semantic Similarity Functions and Their Applications

Similarity is a rich concept deeply rooted in human knowledge and perception. Interest in similarity and categorization of objects can be traced back to Plato. Although studied by philosophers, and mathematicians for a long time, there was no agreement on the “best way” to define it and measure it. Recently, the concept of similarity and methods to assess similarity between objects have assumed great importance in Data Mining (DM), Machine Learning (ML), and Bioinformatics (BI). The various proposed methods to measure semantic similarity do not use semantics and fully agree with human judgement. In this paper we construct semantic similarity functions that remedy this situation.