The problem you've described is actually simpler than the 'classic' recommendation problem, which is personalized per user. All you want is a list of most-similar items. That's a lot easier. You could easily roll your own by using an ItemSimilarity implementation and iterating over all items. No Recommender needed. But actually the item-based recommender class has a "mostSimilarItems()" method which will just do it for you.
On Tue, Jul 26, 2011 at 10:50 AM, Antony Corfield [awc] <[email protected]>wrote: > I've been testing the Mahout Recommender software using a dataModel derived > from activity data generated by page views and downloads of items in an > open-access repository. The taste data has preferences based on the number > of times a user has viewed an item and I've also tested with boolean > preferences (page viewed or not viewed). A user is identified by the request > IP address. > > I would like to base recommended items on the item being viewed rather than > the user but this doesn't seem possible even when using the Item based > recommender. Is this possible or have I missed the point? > > > Thanks, > Antony > -- > Antony Corfield > Project developer > AEIOU Project > Tel. 01970 628724 > http://aeiouproject.blogspot.com/ > >
