In the chapter example, we used a very small sample taken from two Usenet newsgroups (autos and electronics). Use the much larger data set that includes many postings from each of the two groups and again use XLMiner^© to attempt to classify the groups as either “auto”-related or “electronics”-related.

First, text mine the postings and then use the SVD-derived concept document matrix” from XLMiner^© to estimate a classification model that you believe will predict each posting’s category correctly (as either “autos” or “electronics”). Some classification models may work better than others.

Evaluate the resulting classification model(s) using the standard diagnostic statistics and explain their meaning (C11P7).