got interested in spam filters and recommendation algorithms so gave fruition to this classifier.
it evaluates text inputs from "category a" (not really a category but more so a collection of inputs) and discerns whether they're part of category b or not.
this is accomplished through the usage of murmur's hashing algorithm for storage of words as indexes within vectors that then utilize tf-idf in order to state the importance of each key word in every category to create a new vector of weights which is then compared to an "average" (known as centroid) vector through cosine similarity to state whether the given input is part or not of category b