SemTech Thoughts

This post was originally published on the Kasabi product blog.

Attending SemTech 2011 last week I was struck by a shift in emphasis from “What If?” to “Here’s How”. I think there were more people sharing their experiences, technical & business approaches, and general war stories than on previous years. I think this reflects both the extent to which semantic technologies are, slowly, percolating into the mainstream, and the number of organizations that have jumped in to explore what benefits the technology might bring.

Attendance numbers at SemTech remain high, with around 1500 people visiting the conference this year. SemTech has one of the most punishing schedules of any conference I’ve attended, with 9 parallel tracks on some days! This year I changed my own strategy to spend a little more time in the “hallway track”, which gave me plenty of time to catch up with a number of people.

I did catch a number of talks, and while I won’t attempt to review them all here, I will mention a few stand-out sessions. John O’Donovan’s talk on the experiences of the BBC with semantic web technology was the best keynote. I’ve previously seen other speakers from the BBC talk about the domain modelling approach that is yielding great results for them when building websites, but John was able to put some business and architectural contrast around that which I found interesting. I saw echoes of that during the rest of the conference, with the three part architecture — triple store; CMS; search engine — appearing in a number of talks, e.g. from O’Reilly and Entagen. Not surprising as it allows each component to do what it does best, and its an approach I’ve personally used in the past.

The utility of separate search indexes to complement structured queries using SPARQL is something we’re supporting in Kasabi by having both of these options as part of our standard set of APIs.

I also sat in on Lin Clark’s tutorial on using the new semweb features of Drupal 7. We’re using Drupal in Kasabi currently, but haven’t started using these features as yet. Lin gave a great run down of the current Drupal support for publishing and consuming RDF and Linked Data, and I was impressed with the general capabilities.

My main reason for attending SemTech was to give two talks about Kasabi. My first talk was on some of the work we’ve been doing around building APIs over RDF and Linked Data. Our goal is to make make data as useful in as many different contexts and by as many different developers as possible. You can find the slides for these on Slideshare and I’ve embedded them below:

My second talk was a product demo of Kasabi. We launched Kasabi into public beta a few days before SemTech began and I was very pleased to have hit that milestone, allowing me to give a live demo of the product during the talk. I gave a walk through of the site, showing what we’re doing to make datasets more accessible, the ease of publishing both dataset and APIs, and how to quickly import data from the web using a simple browser plugin.

Again, the slides are up on slideshare, and embedded below, but I’m working on some screencasts that should capture the demonstration which was the bulk of the talk.

We had some fantastic reactions to the demo, and lots of interest in the product in general during the event. I was pleased to see Kasabi getting a mention in four other talks during the week. It’s exciting to be able to show more people what we’re building.

I’m looking forward to the new SemTech events later this year in both London and Washington. However Kasabi isn’t just for semantic web developers and so we’ll also be casting a wider net to reach out to developers from a number of different communities.

Attending Strataconf earlier this year confirmed for me that it will quickly become another key event for those of us interested in data. There seems to be a great community forming around the conference already. I did come away from the January conference wishing there had been more discussion of publishing data to the web, rather than simply using data from the web, but I think the emphasis was right for that first event. I’ll interested to see how Edd Dumbill is planning to add a little more semantic web flavour to the agenda of later events.