Hello guys, I used a lot of Mahout especially Taste in my Master Thesis: "An architecture for evaluating recommender systems in real world scenarios". I wanted to give some feedback about it. If somebody is interested in the whole work (97 pages) drop me an email.
First I have to thank you for such a great software product and the fast feedback for tickets that I opened during my thesis (https://issues.apache.org/jira/browse/MAHOUT-650 https://issues.apache.org/jira/browse/MAHOUT-611). It did a great job for me and saved me a lot of time. I was able to get Taste working in an Java EE environment and I implemented my own DataModels. Like a lot of other people I have to admit that getting started is not that easy and it is necessary to make a deep dive into the internals of Mahout. One pitfall for me was that the different IDs have to be sorted. I know it is written in the DataModel documentation :-) I was especially difficult to get the IDMigrator working. Would be quite cool if there would be a DataModel which automatically includes String migration. I had some problems that some interfaces did not implement the Serializable interface. I already opened a ticket MAHOUT-650. Is there a benchmark engine telling RMSE of the different algorithms? Would be cool if a maven command would be available. So when I implement a new recommender I can directly benchmark it against the other implementations. Then I made the following notes during browsing the Taste Source: * getNumUsersWithPreferenceFor for the MySQL DataModel only works for at most two things and there is no warning if more are supplied * DataModel expects that there is always only one rating from a user to an item (what about reratings?) I also attached some images which should explain how Taste is doing it's job in my system. All in all Mahout is a great software project with great people and I am excited that I am allowed to follow the development and when I have code which also makes sense for other people I will create a patch. Have a great week Manuel -- Manuel Blechschmidt Hasso-Plattner-Institut Mobil: 0173/6322621 Twitter: http://twitter.com/Manuel_B
