Ex Libris Recommender January 23, 2009
Posted by Andre Vellino in CISTI, Collaborative filtering, Data Mining, General, Recommender, Recommender service.add a comment
It’s gratifying to learn from Richard that the library product and solutions vendor Ex Libris has developed a recommender for scholarly articles named “bX”. The announcement says:
bX is the result of years of collaboration and research conducted by leading researchers Johan Bollen and Herbert Van de Sompel from the Los Alamos National Laboratory.
Which bodes well for its credentials! bX data-mines log files from link resolvers:
The bX service derives its recommendations from the analysis of tens of millions of transactions performed by users from research institutions worldwide and captured through a large-scale aggregation of link resolver usage logs.
So it probably uses only Collaborative Filtering rather than a hybrid (content-based) scheme. It probably doesn’t use citations either.
Not having seen a link resolver log, I don’t know if there’s user-identity information to be found there beyond the IP address. If there isn’t that could muddy the waters somewhat. Librarians (with the same IP address) often act on behalf of a wide variety of end-users which could lead to rather meaningless serendipity in the recommendations.
The blog for future announcements is here: http://bxbeta.blogspot.com/
Synthese Recommender December 16, 2008
Posted by Andre Vellino in CISTI, Collaborative filtering, Digital library, General, Recommender.4 comments
I have finally finished a first version of the Synthese Recommender for journal articles. It is now up on the CISTI Lab web site, complete with a flash video tour, in lieu of documentation.
For the recommender experts among you – there’s nothing fundamentally new here that you don’t know about already: the Synthese Recommender applies user-based collaborative filtering (implemented using Taste) with article citations as a substitute for user-data to address the cold-start problem. This has been done before in TechLens. And, I should add, TechLens (now in its third iteration) is quite a bit more full-featured and polished.
My aim was modest: to gather data about how well a simple collaborative filtering recommends articles to researchers in diverse scientific fields. That will give me a baseline from which to repeat the experiment – with a content-based recommender, multi-dimensional ratings, and an explanation feature. I’m hoping this will tell me how much more valuable each element is to the overall user-experience. My hypthesis: a hockey-stick curve in usefulness as more features are added. Explanations, I think, are going to make all the difference.
Cloud Computing – is it a Good Thing? February 14, 2008
Posted by Andre Vellino in General, Uncategorized.8 comments
One suspects a computing trend may be passé by the time it hits the mainstream media. Cloud Computing was recently the centerpiece of a CBC Radio broadcast and TV Ontario program and both featured Nicholas Carr, author of The Big Switch.
I am of two minds about the cloud. I like Carr’s analogy with the power grid. The primary advantage of AC current over DC is its ability to transport electrical energy over large distances and to permit centralized generation. People used to generate their own (DC) power, but economies of scale make it more efficient to have it as a centralized utility. Ditto, with CPU and storage. All you need is bandwidth (although, as Daniel points out, the real problem is latency.)
My question is: do you (we) really want to rent centralized commodity computing? Every now and then I advocate (tongue partly in cheek) a return to the IBM mainframe Time Sharing Option (TSO.) After all, who wants to have to upgrade processors and software, or maintain equipment, perform backups, run virus and spam checks? I don’t really enjoy being the sys-admin for the 4 computers (with 4 different operating systems) and the network in my own home: I’d rather be hacking or blogging (or skating, for that matter.)
On the other hand, I also like to excercise unfettered control over my computing environment. I like to know what my processor is doing, I like the ability to unplug from the network, I like to chose which upgrades I do and don’t install. Sometimes I even want obsolete versions of applications (e.g. Yahoo messenger, when it didn’t have add-push and self-updating.) And I’d rather not be dependent on the matrix if I don’t have to be.
I also worry about renting CPU cycles because someone else has control over them. Californians who leased GM’s electric vehicle EV1 were chagrined to discover that the terms of their lease enabled GM to take them off the road (nicely documented in Who Killed the Electric Car.)
As environmentalists now urge us to generate our own power, get off the grid and become self-sufficient, so computing environmentalists will urge users to keep ownership of their hardware and software. Hardware and software don’t have to be obsolete as quickly as they typically become and the longer we keep our computing devices functional, the better (see this recent National Geographic article on high-tech trash being dumped in the third world.)
I think George Gilder was premature to declare the death of the desktop. When people at large discover that WiFi EMF is bad for your health, I’m willing to bet that will trigger a swing of the pendulum away from always-networked, application-free thin-clients and there will be a movement to reclaim ownership of the personal computer in the name of autonomy and freedom.
Synthèse 