I am a current Computer Science student and even with my major, I must agree that most classes tended to waste time when we would use computers in high school. Most of the softer science teachers have kids use computers to make "Powerpoints" and "Videos" and waste a great deal of time doing fun, but generally useless stuff when we could have been learning actual history or English in a class discussion or lecture. I found the teachers that mostly avoided computers (besides the computer science teachers) were the teachers I tended to learn the most from.
However, I still think computers are needed in schools especially in a society where nearly every white collar job requires the ability to use a computer. Also, computer classes, and similar computer-centric classes obviously are going to require a computer lab (at least). I also cannot even imagine how horrific it would be to have to use a typewriter to write all my papers...it's a shuddering thought. Perhaps an emphasis on learning the necessary skills for using a computer in a real life job, rather than an emphasis on integrating computers with existing teaching techniques would create a much more efficient system, while still preparing students for the job world.
I would start by filtering out everything that you know for sure is legal. For me, that would mean all vorbis files,
Next, I would filter out by comments tags. Many distributors like Amazon include a non-drm comment in the ID3 tag. Filter all those out or subtract from their likelihood. If you ever include your own comments in your files, filter those as well.
Now you've gotten most of the obviously legal files out of the way. The next part will be to filter out the likely pirated music from the rest. The user would have the discretion of choosing his/her tests as they see fit. For this, I would probably increment the likelihood variable by one for each matched test.
Example Tests:
1. Low bitrate? (128kbs) to filter out most of the old crappy pirated downloads.
2. Lowercase ID3 tags? Pirated downloads often have typos.
3. Missing artwork?
4. Sketchy ID3 comments (t0rr3nted by r1ppErZ, demonoid, lots of things you could search for)
5. Missing ID3 tags (no album name, unknown artist, etc.)
6. Matching filename with ID3 tags. Often people fix their pirated tags, but the misspelled filename stays the same.
7. Subtract likelihood if bitrate is what you normally rip at
8. Song in incorrect folder. I normally put my tracks in a directory structure of artist/album. You could test the ID3 tags to the directory structure.
9. Interface with MusicBrainz and scan songs for correct tags, correct song length, and if MusicBrainz can even find it.
By no means is that all the tests you could do. Next you would list all the files high to low in a nice format for the user so that he/she could easily spot the pirated music. This solution is obviously quite preliminary and could definitely be refined, but you get the basic idea. The whole point is to make the user's task of finding and marking the pirated software much easier (and I'm sure in easily less than an hour you could find all your pirated music with such a program).
Does the Zune support iPods?
What are you smoking?
He has not acquired a fortune; the fortune has acquired him. -- Bion