Volume is really only valuable when compared to some other volume, either from a historical value, or from some other stock.
The article you linked to doesn't provide specific numbers for you to evaluate whether volume is high or low. Many people simply look at the charts and use a gut feel for whether a day's volume is "high" or "low" in their estimation. Typically, if a day's volume is not significantly taller than the usual volume, you wouldn't call it high. The same goes for low volume.
If you want a more quantitative approach, a simple approach would be to use the normal distribution statistics: Calculate the mean volume and the standard deviation. Anything outside of 1.5 to 2.0 standard deviations (either high or low) could be significant in your analysis. You'll need to pick your own numbers (1.5 or 2.0 are just numbers I pulled out of thin air.)
It's hard to read anything specific into volume, since for every seller, there's a buyer, and each has their reasons for doing so. The article you link to has some good examples of using volume as a basis for strengthening conclusions drawn using other factors.