The Normalized Similarity Metric and Its Applications

Abstract

Similarity metric is important in many applications. In some applications, it is desirable to normalize the similarity metric so as to yield a more meaningful insight of the situation, for example, in comparison of long genomic sequences and comparative gene prediction. In extending the content of a previous work involving formal definition of the concept known as similarity, we give new formulas for normalizing similarity metrics which satisfy the formal definition of the normalized similarity metric. We also describe how these formulas may be utilized in an appropriate application domain in which the use of normalized similarity metric is desirable.