September 2017
Beginner to intermediate
412 pages
8h 55m
English
If we think of each column y of the utility matrix as an n-dimensional vector, y = (y1, y2, ..., yn), then we can use the Euclidean dot product (inner product) formula to compute the cosine of the angle θ that the two vectors make at the origin:

This is called the cosine similarity measure:
For example, if y = (2, 1, 3) and z = (1, 3, 2), then:

We can see that the cosine similarity measure has the six requisite properties for a similarity measure. If u and v are parallel, then s(y, z) = cos θ = cos 0 = 1. That would be the result in ...
Read now
Unlock full access