Search Engine Meeting in Den Haag
I've spent the last couple of days in Den Haag in the Netherlands, at the Infonortics "Search Engine Meeting". Whilst it isn't what I currently do at the BBC, a presentation proposal I submitted some time ago was accepted, which I gave today. The meeting was sponsored / co-hosted by the European Patent Office, and we had a reception at their offices last night.
A few things that struck me:
Business taking Search more seriously Now the big players have consolidated there seems to be increasing activity around the fringe of the "core" search market - with smaller companies making innovative products that are stealing big contracts from the obvious candidates. I was particularly impressed with the thinking of Mondosoft. There also seemed to be a lot more mileage in going to businesses with intranet or data mining solutions that took a more holistic approach to structured and unstructured corporate data, rather than just trying to sell "search-in-a-box".
English / Non-English Language Split There was an interesting panel of nearly two hours talking about machine driven multi-lingual translation and IR software. The first question was from an American: "But is there any need for this?". The clear implication was that the 12% (or whatever) of Dutch people who can't speak English either had nothing to offer the rest of the webiverse, or that the webiverse had nothing to offer them if they were too lazy to learn English. Joop van Gent's retort was excellent: "Let me ask you a question back. Was there ever any *need* to make the programme Big Brother? Hmm? No, there was no *need*, but it turned out to be a commercial opportunity that people enjoyed and made money". I'm not sure about using the example of Big Brother as a rallying call to do something, but it really underlined a strong impression that in Europe there is a feeling that the historical English language dominance of the web should be turned into an early adoption phase, rather than become a fait accompli.
Defense [sic] Applications On Show Yesterday we saw screen-shots of applications that could monitor Control & Command Crisis Management Information visually in real-time, we saw a search engine used by American Intelligence to deal with terrorism, we saw a dynamic metadata classification tool that could identify gaps in intelligence information on a geographical basis. We also saw a long presentation about the kind of linguistic analysis tools used by the United States Government, and the image-matching tools they use as well. Search and Information Retreival systems are becoming big business in the defence and intelligence industries.
Other things to note:
- I finally heard someone casually use the word 'maven' in conversation.
- I didn't realise computational linguists could get so annoyed with each other.
- A fire alarm went off as one of the speakers was showing a screenshot of some US Government Intelligence admin systems - everyone freaked, we thought we had been busted!
- I got to talk to Karen Spärck Jones about EastEnders.
Mostly though, I still haven't got over the fact that the approved hotel in Den Haag on the BBC's booking system is called Badhotel. Thankfully it has in no way lived up (or should that be down?) to its name. Notably they have 30 minutes free internet access in the bar, provided you buy a drink. This did not prove to be an obstacle.