I have gone through it carefully, and it is clever.
OTS is definitely word freq based.
2,303 karma · joined May 9, 2009
I have gone through it carefully, and it is clever.
OTS is definitely word freq based.
I'm watching this thread with interest because I've been doing some work on building a good summarization engine for a few months now, and it is getting pretty close to ready for prime-time -> summarity.com
Only if you can provide some ex-ante assurance to the good ones that your reputation system isn't going to be gamed by bad ones who focus on gaming rather than being good recruiters.
No hard feelings, sorry if I was reading too much into it.
I have search low and high for a usable automatic text summarization library, but everything was a high-dollar item for finance
Now you really have my attention. I'll shoot you an email.
The coding is all done in python. I'm using nltk, although the approach is more statistical inference than linguistics. That's really the only library. I'm very much a bottoms-up guy, and am happy to re-invent the wheel because it sometimes gives me a better understanding of the texture of the problem I'm up against. For example, where I'm doing some linear algebra / graph traversal math I'm using functions I've written rather than numpy, or scipy, or a graphdb, etc.
It won't go into a lot of detail, but it should be a good overview. The truth is that I still have lots of ideas on how to make it better, but it is also the case that for every 8 things tried maybe 1 works. It comes back to your point that part of the appeal of working on this class of problems is that it is somewhat hard. In general, I would prefer to work on things that have technical barriers to entry, rather than network effects.
The reason I ask is because I know I'm not going to watch 20 hours of the videos, and when I take a look at the transcripts they are good, but they are also long. It would take a while to even skim them.
I've been doing some work around machine text summarization, and it seems like these transcripts would be a great candidate to put it to work.
What do people think? Would you like summaries of the transcripts of these lectures?
EDIT: Thanks for the positive feedback, I'll submit them as a story here in 24 hours (right now I'm supposed to be getting ready for a big dinner tonight).
Their tagline is 'making your interests special interests.' Their investors include Founders Fund, Conway, Dixon, McClure, Rabois, etc. And they are pretty clear that what they're pursuing is an 'unmatched opportunity to change representative democracy' (i.e. aiming high).
It pays to be an optimist.
It's that he has one of the best investment processes.
The only place I've really seen the one year clock is just as something universities made up to keep their tuition price discrimination effective.
"The only answer for the developed world is technology and innovation. We can't compete with China on labor costs, this is not an option. And globalization without technology means an increasingly Malthusian struggle for resources."
Personally I find this to be one of the best endorsements of YC that I've read. How people treat you when things don't work out is really, really important.
Good job, Tam.
Find a situation where there is a guy who is looking for a solution to his problem at least as hard as you're looking for a customer for the product. Don't try to convince a bunch of people who don't seem to want to be convinced, that is a really painful process that leaves you disillusioned and makes you wonder if you should quit. Find one of the few guys who is going to love you when he sees your stuff. And if your stuff works for him, he'll probably help you expand.
Bottom line: spend more time thinking about applications and finding people, and less time trying to convince them
Are there startups that provide this sort of fraud detection, the way Directed Edge does recommendations?