Clustering newsgroups data using k-means

The newsgroups data comes with labels, the categories of the newsgroups, and a number of categories that are closely related or even overlapping, for instance, the five computer groups: comp.graphics, comp.os.ms-windows.misc, comp.sys.ibm.pc.hardware, comp.sys.mac.hardware, and comp.windows.x, and the two religion-related ones, alt.atheism and talk.religion.misc.

Assuming those labels do not exist, will samples from related topics be clustered together? We, herein, resort to the k-means clustering algorithm.

How does k-means clustering work?

The goal of the k-means algorithm is to partition the data into k groups based on feature similarities. K is a predefined property of a k-means...