Sunday, June 03, 2007


The end of the day was devoted to company time. I went into Indy's talk about fourthcodex products, in order to give him my moral support.

A key component of the software from fourthcodex is Codex (called OntoDLV until a couple of days before the conference), which provides a number of unusual (and useful!) processing facilities, and is quite different from OWL. Indy had been jokingly saying that he would announce that everyone has it wrong with using open world reasoning (the basis for OWL). I wasn't sure how much of this was humor, so I suggested toning down the "you're all WRONG" angle, and instead point out how the strengths and weaknesses of closed and open world reasoning. After all, saying that the OWL approach is wrong seemed like a courageous statement at a conference where everyone else was touting the benefits of OWL.

Maybe I went too far when I suggested toning it down, because when Indy got to the point of pointing out the weaknesses in OWL, his body language got all defensive, and it took him a minute to make the point. It's hard enough getting up in front of a room full of strangers to explain the advantages of your product, so I felt pretty bad.

I'd have loved to attend Tom Ilube's presentation from Garlik during the following session, but instead I got word that we were all asked to get together for a last minute preparation before the Exhibition session.

In the Exhibit Hall

The exhibits had lot of companies (mostly new companies) talking to potential clients and checking out the competition. This was also the occasion when food and drinks were being served. I would have loved some, but since we were working we couldn't have any.

There was a real buzz around our booth. I don't know how long it really was, but it felt like I spoke continuously for 2 hours. It was very draining. However, most of the people I spoke to seemed genuinely interested in our software. After all, some of it is pretty interesting. There are other closed world reasoners (any SQL database can do that, and indeed Codex stores information in an SQL database), but disjunctive reasoning is quite unique, and offers some interesting capabilities. Supporting non-monotonic and some defeasible reasoning is also unusual, and very useful. With the exception of Alex, the other products are then built on this engine, so they each have some unusual and useful features as well.

I think the products are interesting, but this is the first time that I've seen potential customers understanding the benefits of features like this. It ended up being an interesting night, and I came away feeling better about what we had to offer than when I went in. Will wonders never cease.

While it was all wrapping up, I heard that there was a panel discussion going on. IT wasn't on the program, but it included a few notable names, and I was very interested to hear it. Unfortunately, work kept me there much later than I'd expected, so I missed a lot of it. All the same, I got to hear some interesting perspectives from both Eric Miller (a semantic web proponent) and Ramanathan Guha, now from Google.

When I first walked in, it sounded like Ramanathan was saying that a semantic web won't work, which was leading to some lively discussions with Eric. Eventually Ramanathan explained that he doesn't object to the semantic web at all, but rather he doesn't believe in semantic search (at least, that's what I think I remember him saying). Semantic search requires automated semantic extraction, and Ramanathan made it clear that it is Google's opinion and his that such a thing is not feasible - at least, not in the foreseeable future.

Unfortunately, it all finished only a short time after my late arrival. So then it was off to dinner with clients, giving me a second opportunity to go outside since arriving. I really wasn't seeing much of San Jose at this rate! Maybe I should go back and visit, say in the Chicago winter.

No comments: