Weblogs

January 8, 2008

10:43
So Identity Theft is not a crime but fraud is. Maybe identity theft should be means tested, so if you steal from someone poorer than you THAT would be a crime, but stealing £500 from Jeremy Clarkson would be fine… or even an encouraged form of wealth distribution.
Categories: Weblogs
04:15
I think the fact that CSS frameworks exist is proof that they are too complicated. Yet Another Multicolumn Layout | An (X)HTML/CSS Framework. I love the idea though, that IF WE CAN JUST AGREE ON WHAT THE XHTML ELEMENTS ARE FOR GOD’S SAKE… that the rest falls into place. A Plone skin would work on [...]
Categories: Weblogs

May 4, 2006

11:27
This isn't advice for you to personally pursue some negative experience. It's to find people embroiled in negative experiences -...
Categories: Weblogs

March 9, 2006

08:49
"The one thing that we know for certain is that context is vitally important; it will shape the opportunities in...
Categories: Weblogs

March 7, 2006

08:11
In recognition of the passing of Dana Reeve, I was watching the Today Show dedicatory coverage. It that piece, Christopher...
Categories: Weblogs

February 14, 2006

20:46
All but one thing can motivate me enough to make a post to this site when I'm just otherwise...
Categories: Weblogs

October 13, 2005

10:15
I forgot to mention it at the time (Damian didn't), but we recently hacked up a bbc backstage browser as a fun way to mess around with open data, in this case TV and Radio Listings. (I should say many others have also done this but afaik we're the on...
Categories: Weblogs

September 30, 2005

05:50
Here's an interesting paper on folksonomy from Scott Golder & Bernardo Huberman at HP Labs (although they eschew the f-word and prefer the term 'collaborative tagging'). It's basically a statistical analysis on del.icio.us bookmarks (a 72 hour ti...
Categories: Weblogs

August 3, 2005

13:46
One hoary old example we tend to use in the semantic web is "the difference between silicon chips and potato chips'. How would you do this on google? Well you'd probably add disambiguating words (like, er potato). Doing that sort of thing with photos l...
Categories: Weblogs

August 2, 2005

06:50
One thing I haven't talked about so much is the facet that we are taking the ideas laid out my my ACM semantic blogging paper and applying them inside HP. To do this, I have recruited Paolo Castagna (one of the creators of the PlatypusWiki) to HP Labs ...
Categories: Weblogs

June 30, 2005

14:00
It’s time for another round-up on CNet of what Google, Microsoft, Yahoo and others have been up to recently.
Source: unstruct
Categories: Weblogs
13:33
The June 2005 issue of the ACM SIGKDD Explorations is a special issue on Natural Language Processing and Text Mining. Read full text at http://www.acm.org/sigs/sigkdd/explorations/issue.php?issue=current Text Mining and Natural Language Processing Introduction for the Special Issue Anne Kao, Steve Poteet Mining Knowledge from Text Using Information Extraction Raymond J. Mooney, Razvan Bunescu Instance Filtering for Entity Recognition Alfio Massimiliano Gliozzo, Claudio Giuliano, Raffaella [...]
Source: unstruct
Categories: Weblogs

June 2, 2005

11:45
Yahoo has released a new search service called Mindset that classifies sites as being either commercial or non-commercial, with which they mean informational. The commercial pages are such that want to sell you something. Each page is given a score between -2 and 2 with 0 being a pages that is “balanced”, whatever they mean [...]
Source: unstruct
Categories: Weblogs

May 16, 2005

16:12
Microsoft has made their experimental web search result clustering available at the MSN Beta site. It is very unimpressive as of now with very poor presentation of the results compared to, e.g., Vivisimo, but perhaps they can improve it in the future. Also, it only works in Internet Explorer.
Source: unstruct
Categories: Weblogs
10:52
Microsoft has released a new version of their desktop search. It now runs on Windows 2003 without special tricks and supports a lot of iFilters. A quick glance seemed indicated that they are at least as capable as the new Google desktop search that was released recently.
Source: unstruct
Categories: Weblogs

May 12, 2005

16:03
We’ve written about IntelliSeek’s BlogPulse here before. There is a now a new and improved version that tracks 10.8m blogs growing with 47k per day. Over 400k entries are added each day. They track popular links, people and phrases as before in the blogoshere.
Source: unstruct
Categories: Weblogs

May 7, 2005

12:38
Position Title: Research Scientist Location: Seattle, WA Description: Be part of a mature company conducting artificial intelligence research and development. Work cooperatively to conduct and apply cutting edge research, write research proposals, architect and design systems, direct small teams of developers and participate in implementation, testing and integration. Stottler Henke’s Seattle Office specializes in developing [...]
Source: unstruct
Categories: Weblogs

May 5, 2005

14:52
Fortune has an interesting inside look at Microsoft’s problems in trying to beat Google in the search game.
Source: unstruct
Categories: Weblogs

April 21, 2005

11:43
Job title: Research Engineer Location: San Diego, California Description: You will work with a small team of scientists creating innovative solutions to customer problems. These new solutions often form the basis of new or improved products in the Kofax family. Responsibilities include implementing and designing algorithms for natural language and image processing (e.g. classification, extraction), prototype system development, [...]
Source: unstruct
Categories: Weblogs

April 13, 2005

09:16
Not long after their last price reduction, Google has again reduced the prices on their appliances. This is interested since it clearly shows where the market is going, and fast. Now you can index 100k documents for $3k, a price reduction down to a third. The higher-end boxes now start at $30k for 500k documents [...]
Source: unstruct
Categories: Weblogs