A couple of weeks ago we reported on the results of the latest Elsevier Semantic Web Challenge (see story here). That led us to wonder: What’s been going on with some of last year’s Open Track winners (covered in this story here)?
We’ll start with TrialX, which is turning semantic technology and social media into an online clinical trial patient recruitment business. Last year we were impressed by the service's grasp of how important it is to present a user-friendly front to consumers, who can visit the site and enter information about their health condition and then use its decision engine to map up to clinical trials that may be appropriate for them to participate in, or even use their personal electronic health records to find out about trials.
The business model, of course, is about getting the recruitment companies on board as a lower-cost alternative to non-targeted print and radio ads: It may be doing something right on that end, as it reported in November that, with 170,000 visits a month as counted up by Quantcast, it’s the highest visited site among patient recruitment companies.
November also saw the outfit present research at the 2010 American Medical Informatics Conference that took its web log data (anonymously) and semantically analyzed the aggregated information to better understand what patients actually are searching for in clinical trials and what terms they use to phrase their queries. The idea behind that is that having better insight into this can help close the gap between what patients are looking for and what information is available.
By dynamically tailoring the types of information presented to the user using semantic pattern analysis over server query logs, TrialX believes it can contribute to reaching an even more important end: The delivery of the right information to those consumers will encourage their enrollment in clinical trials – trials which often get delayed for lack of eligible participants. Its results showed that highlighting information nuggets around what’s relevant to those searchers – a combination of condition, location and medication/treatment – may be an important way of keeping those potential participants on sites long enough to discover trial participation options.
TrialX appears to be doing well by doing some good -- as famously said by CEO Bigweld of Bigweld Industries in the animated film Robots, "See a need, fill a need."
Another of last year’s winners was VisiNav, which lets you search and navigate web data from multiple sources into one spot. But it looks like the last update of objects and statements to that system was in November of 2009. In comparison, the other winner in the group, Sig.ma, the search browser, mash-up generator and API for the web of data that also hails from the Digital Enterprise Research Institute (DERI), put out a version update at the end of last month after a slew of them throughout the year.
Sindice is the power behind Sig.ma, and the minds behind the efforts noted on their blog this month that an infrastructure upgrade is on the way. Expectations are for it to increase both the capacity of individual servers and introduce failover support so that, over the next few months, it can significantly increase “the amount of data it can both fetch from the web of data and, in turn, process, transform and publish back to the semantic web community.”
OK, so a few trial runs do show that it does take a bit of time to retrieve all the data, but what a picture of the web of RDF, RDFa or microformat entity data extracted from multiple web sites that you can get from it. You can spend hours wandering along the (sometimes noisy, as the service's creators admit) primrose path, and word is that it’s pretty easy to let Sig.ma know about your own semantic data and get it into its search results.
As we mentioned when Sig.ma won its award, it would be nice to work on the user-friendliness (something for which this year's entries overall received kudos from the judges). For instance, “Solo” as a menu button doesn’t seem like the best word to use to indicate to users that clicking on it is a way to show data only from a particular source.
What progress will this year’s Elsevier winners have come this time next year? Stay tuned…