Recommender Algorithm

Question

I need to put together a Recommender Algorithm for a website. I have come up with a very simple method of achieving this but was wondering if anyone could point me towards any literature or such that could help me get a better idea of how other examples have been put together.

I have been made aware of functionality such as Collaborative filtering, Clustering and Categorization in Apache Mahout but am not to clued up about how the Machine Learning fits into all of this. I can see how to make algorithms for the above (apart from Machine Learning) but was wondering if anyone knew of anything else which could be added to the mix.

Also, what would you say the purpose of a Recommender is, how might it best function? Anyone willing to share a definition?

Thanks!

Answer 1

There is an article which discuses the different possibilities of putting together different algorithms and creating a recommender. The authors have analyzed 37 different systems and their references, and have sorted them into a list of 8 basic dimensions.

Although the paper has been published on 2003 and some of its examples aren't available now, still it can be a very good starting point for researchers to construct their own recommender system.

I'd like to share Robin Burk's definition of recommender systems as in his paper:

Any system that produces individual recommendations as output, or has the effect of guiding the user in a personalized way to interesting or useful objects in a large space of possible options.