Dunno, I think the raw frequencies work fine in your case, because there aren't really any themes that all (or many) authors across your selections keep returning to.
The exact details are out of my hands -- I'm just a tech guy -- but we're active members of the music information retrieval community and always have been.
No-one even knows if it's possible to crowdsource good enough BPM data like this yet, so even demonstrating that it's feasible would be progress :-)
This means you can't use any interesting characteristics of the music itself, or the associated metadata, to aid the recommendations. All the interesting domain knowledge is stripped out, which likely means the best solutions still won't work as well as algorithms that use metadata (like Last.fm's) or content analysis (like Pandora's) or both, and certainly won't lead to any particularly interesting insights about what drives people's tastes.
Yes, and graduates with an interdisciplinary background and strong reasoning skills are more likely to get the interesting jobs than pure software engineers who know JUnit inside-out.
(Generalizing from myself with a sample size of one)
As a counterpoint, I like this quote from Twitter's Nick Kallen:
This smacks of the oft-ridiculed Java AbstractFactoryFactoryInterface. But let me put it bluntly: AbstractFactoryFactoryInterface's are how you write real, modular software–not little fart applications.
[N.B. I'm not saying there isn't a lot of truth in the factorial article, it's just you have to know which challenges just need a one-liner function and which require an AbstractFactoryFactoryInterface]
It may be a 'great' name ideologically, but the fact that there are three other comments in the thread giving three different ways it's pronounced, shows a certain degree of name fail.
EDIT: Sorry, five different pronunciation suggestions at last count.
Sadly, superficial things like names are important if you want to compete with better-known products.
I can't even pronounce LibreOffice fluidly -- there are no words in English (I think) with a schwa followed immediately by a short 'o' sound, so no native English speaker is phonologically equipped to deal with it.
That's terrible advice, I hope you're being sarcastic but I fear not. There's plenty of useful stuff outside of CS which isn't liberal arts.
Maths, stats, electronics, physics could all be useful in an entirely computing-based career.
Biology or chemistry could open up a career in bioinformatics, molecular modelling or simulations. Likewise linguistics for text mining, information retrieval, speech/language processing.
Economics if you're interested in being an entrepreneur.
The crunchier end of philosophy, where it overlaps with maths and linguistics and cognitive science, will give you a much deeper frame of reference for understanding many hard problems.
Not all computing jobs involve twee social web startups or mundane CRUD.
If I could go back and do it all again, I'd definitely do more stats courses.
Also if you're interested in data science in general, some basic linguistics (syntax/semantics) would be useful. (Saying this as someone with a PhD in natural language processing, who had to self-learn all the linguistic background the hard way)
"It is fundamentally different from the ad platform that is Google. People go to Google to find something they need, possibly ready to buy, which a good percentage of the time can in fact be solved by someone's ad. Facebook ads, on the other hand, annoy users. They yield no real value, and thus no profits. "
Err -- television ads also just serve to get in the way and annoy users, when they want to sit and relax and do something completely different from hunting-for-stuff-to-buy.
But last time I checked, most TV channels are still running ads, 50+ years on.
But, thanks :-)