What would be interesting as well [but is certainly impossible for me to do] would be to do the reverse of "Data Mining" as available for example in ShareScope. In ShareScope one sets a number of different parameters such as Capitalisation, PEG, Yield, etc., etc., and the software will produce a list of companies that meet those criteria.
It would be good to "reverse engineer" this by devising a software programme which could analyse the companies on our "list" and produce a list of what their common parameters are!!
