Calculating similarity between two data items is a fundamental task in data analytics, machine learning, and data science.

• Similarity is a real-valued score describing how similar two data items are.
• It is an essential step before many other tasks (clustering, classification, retrieval).
• First, convert each data item into a feature vector (e.g., an employee → <M, 51, 64000>; a document → its term-frequency vector).