Comments for Timboy

Comment on Hiring: false positives and negatives by SKMurphy » Plus Minus People

SKMurphy » Plus Minus People — Tue, 06 Jan 2009 22:23:28 +0000

[...] team over someone who has unanimous but lukewarm support. Tim Converse wrote a good post on “Hiring: false positive and negatives” that analyzed categories 2 and 4, noting that allowing for some number of 4’s allow [...]

Comment on Proximity is a hack by Interesting, Cool & Useful - Sep 12 - Dave Naffziger’s Blog

Interesting, Cool & Useful - Sep 12 - Dave Naffziger’s Blog — Wed, 12 Sep 2007 07:00:53 +0000

[...] Proximity is a Hack. Tim Converse, formerly of Yahoo Search and now at Powerset, offers a great writeup about the limits of keyword proximity in determining relevance. [...]

Comment on Proximity is a hack by Negative SEO blog : SEOassassin » “Natural language processing” my natural black ass…

Tue, 31 Jul 2007 12:07:26 +0000

[...] Tim has a thought-provoking post about proximity and it’s uses and limitations, “Proximity is a hack“. I’ll let you read it youself, as it is quite long, but it’s well worth [...]

Comment on Proximity is a hack by Tim Wintle

Tim Wintle — Mon, 16 Jul 2007 15:21:54 +0000

Have to say this is a theme I have heard coming from multiple sources recently, about moving away from pure vector based semantics.

I certainly do not agree with saying that we have squeezed as much as we can out of vector models, although possibly out of pure linear spaces.

I definately agree with the idea of doing further linguistic work to the text first, though, and I thoroughly enjoyed this article. I think the emphasis has to go on combining something along the lines of neural networks with the existing vector models, where many people I have heard talking about this recently have suggested completely removing the vector approach – which seems to be throwing away years of work.

Comment on Proximity is a hack by brendan

brendan — Fri, 06 Jul 2007 03:56:45 +0000

epicycles to epicycles — I like it!

Comment on Proximity is a hack by randfish

randfish — Mon, 02 Jul 2007 16:27:30 +0000

Tim – so, if I understand you correctly, the idea here is to have the machines actually “understand” context and then use it to make better correlations between term proximity, link relationships, etc, yeah?

Thus, what you’d really have is something close to natural language processing, right?

Gotta say – if Powerset can do it before Google does, there’s going to be a lot of very impressed people out there, myself among them.

BTW – Congrats on the move and the position. Haven’t seen you since Chicago, but hopefully you’ll emerge from coding for San Jose

Comment on Working at Powerset by MikeD

MikeD — Mon, 11 Jun 2007 05:54:42 +0000

Yeah, but what if it’s a million CPU years – and you get a million computers working on it?

Comment on Working at Powerset by Jeff Barr’s Blog » Links for Sunday, June 10, 2007

Jeff Barr’s Blog » Links for Sunday, June 10, 2007 — Sun, 10 Jun 2007 17:30:19 +0000

[...] Tim Converse: Working at Powerset – “If all this sounds computationally expensive to try to apply to the entire web, well … it is. There’s a big double bet here, on a couple of decades-long trends: that NLP technology keeps getting ever faster and more mature, and that Moore’s Law continues making computation ever cheaper, and that the historic moment when NLP meets Moore’s Law in the middle at web-scale search is …. 2007. Or maybe 2008.” – Via TroutGirl. [...]

Comment on Working at Powerset by Timboy at Powerset « Fishy thoughts

Timboy at Powerset « Fishy thoughts — Sun, 10 Jun 2007 15:27:33 +0000

[...] 10th, 2007 at 3:27 pm (Silicon Valley) Congratulations to Powerset for snagging Timboy! We came to Silicon Valley more than 7 years ago to do startups, and although the big search [...]

Comment on Working at Powerset by ifindkarma

ifindkarma — Sat, 09 Jun 2007 01:07:22 +0000

I never paid attention to logic in college.

1) What Powerset is trying to do is so hard that it couldn’t be done in a million years.

2) Powerset is lame because it hasn’t launched already.

Ergo:

3) Powerset is lame for trying to do something so hard that it couldn’t be done in a million years.

I guess I could see statements 1) and 2) being made by people who think the only cool companies are the ones tackling easy problems.