Vector similarity is quite similar to the use of bags, however much
easier to use in tandem with the standard math operations we learned
in linear algebra. One of the additional features of vector similarity
over bags is its ability to handle word dependencies, such as oranges
being related to tangerines. We have thus far discussed three types of
comparisons that can be performed using vectors: the cos angle, the
euclidean distance, and the dot product distance. These comparisons
can then be use TF (Term Frequency) and IDF (Inverse Document
Frequency) to give importance to individual words.
