the mr. smith, jones and robinson puzzle was somehow solved by a computer program in the 70s. the designer was paul weston | ralf barkow scanned the drawing and added some more information.
weston worked on it at the bcl lab, aka heinz von foerster lab, in illinois, u. champagne. actually i am missing a reference to that and other works badly, when i read the paper efficient name disambiguation for large-scale databases, which gets highlighted by technews this week.
the core of this work is an algorithm, which should be able to

cluster together names that the system believes to be similar ... ...the system correctly identified authors 90.6 percent of the time, when tested on 3,355 academic papers written by 490 different researchers.

it is said that it will be included in the next version of citeseer, the scientific papers search engine.

"they knew they had to work on meaning at the early days of machine translation", karen sparck says, who actually writes her name spärck, karen spärck jones. often they call her sparck - at least this writing helps within google world.

she calls the semantic web an naive approach driven by computer scientists, who don't know what they are talking about. she doesn't believe that knowledge can be captured by ontologies. universal character (audio/x-mpeg, 618 KB) .
revisited

let's call it semantic blogging

Firstly, we enriched blog metadata using arbitrary vocabularies (this is in fact a key feature of RSS 1.0 [RSS Modules] ). Secondly, we used this metadata to provide navigation and query interfaces onto the blog. Thirdly, we encapsulated information snippets (which we call items) into the blog, such that each blog entry is an annotation for an item. These items carry their own rich metadata which can be navigated over and searched for. Finally, we experimented with using RDF for UI configuration, allowing run time control of the blog look and feel.

there is an online demo and lesson learnt paper. at first i wondered about why they have ....2001/ in their url of 'lesson learnt', i couldn't believe that hp would have said three years ago

We also noted that blogging is a very successful paradigm for lightweight publishing ...

well, depending on the month - blogger came out spring 2001, no wait, that was the year when the first problems appeared because of popularity, it was 1999 - still, they would have been something like early adaptors. ah, at the end of the paper: 2003 ...

One possibility for the future would be to conflate blog content and metadata (there is only metadata; the blog entry content is created directly from the metadata).

hm, if i take this, i have a blog that creates content by using parss and nothing but parss, no?

... The category chooser client goes through the text for a blog entry, stemming words and matching them with the indicator terms. If a term is matched, then that concept is suggested. This is a form of ontology driven assisted metadata creation ...

and pretty soon they will all sound the same. a real killer [app]