jump to navigation

Synthese Recommender December 16, 2008

Posted by Andre Vellino in CISTI, Collaborative filtering, Digital library, General, Recommender.

synthese1I 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.

This is how to use Synthese:

  1. Search an index of about 1.5 Million BioMed articles
  2. Add important articles to a “baskets” of favourite articles 
  3. Recommended other articles based on that basket
  4. Rate how relevant the recommendations are for your research

Users can search on the author / title / keyword / abstract fields (the next version will be on the full-text as well), view the articles from the publisher’s site, create multiple topic-baskets, generate recommendations based only on a given article’s citations and keep a list of previous searches that produced results for each of your baskets.

In previous version, I tried visualizing recommendation results with a prefuse applet, but the initial coments I got back from most people who tried it was that this was worse than useless.  

Although I don’t know a lot about biology or medicine it looks to me like recommendations are in the right ballpark.  By which I mean, they appear to strike some balance between serendipity, diversity and  semantic relevance to the original article basket.

Let me know what you think!


1. The Synthese Recommender System - December 16, 2008

[…] Andre Vellino has just opened his Synthese Recommender System: a recommender for journal articles. Andre works for one of the largest scientific library in the world (CISTI). You can read all about his project on his blog. […]

2. Daniel Lemire - December 16, 2008

Interesting tool. Hope to see it grow. This is an interesting and important problem and we need a diversity of tools.

3. Ricardo Niederberger Cabral - December 16, 2008

Is the source code available? I’ve started a project at http://code.google.com/p/recfwk/ which might be of your interest: a software framework provides basic code for helping the process of designing and implementing recommender engines.

4. Andre Vellino - December 18, 2008

Thanks for your comment, Ricardo. Your project looks very interesting. I look forward to studying it more carefully.

The source for Synthese is not currently available. It’s not “closed” in the sense of “proprietary”, it just isn’t really ready for public consumption. Basically, the core is a set of classes built on top of the Taste framework for handing objects like “articles” and user preferences and citations etc.

I’ll certainly give some thought to making this open source if it’s of general interest.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: