Classification is a key technique to analyzing data that yields very different results based on the classification scheme. One technique is to use two different classification schemes on the same dataset, do the same analysis and compare the results to find hidden insights in data.
As an example, in the 17th century book De Tribus Generibus Instrumentorum (On the three types of musical instruments), the author Francisco Bianchini describes two schemes: the first classfication scheme distinguished instruments by how the sound is made:
- Tensile (String — by plucking)
- Inflatile (Wind — by blowing air in)
- Pulsatile (Percussion — by striking)
He also describes a second classification scheme which was proposed by Augustine (and later by Isidore of Seville) based on the to three divisions of music: harmonic, symphonic, rhythmic:
- vox (harmony — by human voice)
- flatus (symphonic — blowing in)
- pulsus (rhythmic — striking something)
The two classification schemes overlap somewhat (wind instruments are same category in both), but the “pulsus” category includes string instruments and percussion in the second category, but only percussion in the first.
An interesting application would be to use both taxonomies to aggregate data and see what insights can be gained from each grouping.