Hi all and thanks for your suggestions, I am reading the content suggested and, for the "old" (non-spark) approach, trying to see the various options from the javadoc here https://builds.apache.org/job/Mahout-Quality/javadoc/ and the source code.
If I need to use a classical user-based technique, however, the only alternative is the Taste-oriented code, am I right? Still, I can't see how to perform a prediction for a a user/item couple, is there a class for that? Thanks, Eugenio 2015-02-12 17:54 GMT+01:00 Ted Dunning <ted.dunn...@gmail.com>: > I would go so far as to say that all of the old Taste-oriented code is > strongly deprecated. The indicator-based approach that Pat refers to is > the best way forward. > > > On Thu, Feb 12, 2015 at 8:29 AM, Pat Ferrel <p...@occamsmachete.com> wrote: > > > The new cooccurrence recommender that works with a search engine has > > several references at the top of the page here: > > http://mahout.apache.org/users/recommender/intro-cooccurrence-spark.html > > > > On Feb 12, 2015, at 6:56 AM, John Hofmann <genghisu...@gmail.com> wrote: > > > > You're not missing any secret cache of mahout documentation as far as I > > know. I learned what the recommender options were by looking through the > > source code. They're spelled out there. If you know any C-based > languages > > you can navigate the code pretty easily but expect to spend some time > > getting familiar with the repo. > > > > The other avenue I've taken was the book "Mahout in Action." It's a > little > > dated, but is generally still pretty applicable. It goes into more > detail > > about why you'd pick one option over another. I've noticed that most of > > the blog posts about mahout assume you have a level of knowledge about > > different ML algorithms that is comparable to what's in this book, so > it's > > a good one to read if you are going to be doing serious work with Mahout. > > > > There might be better options, but that's how I learned. Hope this > helps! > > > > On Thu, Feb 12, 2015 at 9:24 AM, Eugenio Tacchini < > > eugenio.tacch...@gmail.com> wrote: > > > > > Hi all, > > > I am new to mahout, it has been a couple of days now since I started > > > working with it and I've found it very very powerful. > > > > > > I noticed, however, a general lack of documentation. I am working just > > with > > > the recommender system features and, correct me if I am wrong, I can't > > find > > > anywhere some complete documentation about. All I can find in the > > official > > > website is some quick-start tutorials. > > > > > > Knowing the theory about recommender systems, I expect a few very > simple > > > documentation pages which tell me something like: > > > > > > - The algorithms implemented are: 1) user-based, 2) item-based, 3) > > > .......... > > > - To use 1) in your code, follow these steps: > > > - choose a user similarity measures (available measures: 1) Pearson > > > Correlation 2) Cosine similarity 3) ..... 4).......) > > > - choose a neighbors selection techniques (available techniques 1) > > > threshold 2) fixed number 3).... 4)....) > > > - compute the neighbors using the following > > > UserNeighborhood neighborhood = new ThresholdUserNeighborhood(0.1, > > > similarity, dm);....... > > > > > > and so on until all the options available are covered. > > > > > > I didn't find anything like that, for example at the moment I am trying > > to > > > figure out how to compute a prediction (user-based algorithm), e.g. > > predict > > > the rating for user x movie y but I didn't find anything about that. > > > > > > Forgive me if there is something I am missing, I just want to focus on > > the > > > right content to learn about mahout, any suggestion is welcome. > > > > > > Thanks > > > > > > Regards, > > > > > > Eugenio Tacchini > > > > > > > >