Mahout Content Based Recommendation Engine

2
votes

I am working on a recommendation problem (Content based recommendation). I have my data set in mongodb in json format.

Problem Statement
There are items which have their own properties, and users have some preference regarding each properties. Now I am thinking to predict how much the item x will be liked by the user based on the properties of item and comparing the preferences of the user for same properties that item x have. I want to build a recommendation system to recommend the items to user , based on their preference.

I am thinking of using Mahout and CBAYES Classifier algorithm to predict , "how much item x will be liked by User A ". But I haven't found any example and data set for implementing CBAYES using mahout.

If you have any other suggestion to use any other classifier algorithm then please recommend.

1
algorithmmongodbhadoopmahoutmahout-recommender
What actually troubles me with your question, framework aside of course, is how are you intending to use the Complementary Naive Bayes classifier as a recommender algorithm. Could you explain please? And what is your actual question? So would be helpful too! – eliasah

1 Answers

1
votes

You can calculate “how much item x will be liked by User A” by using cosine similarity. Please refer the following link for your more information.

Reference link: What's difference between Collaborative Filtering Item-based recommendation and Content-based recommendation

Regards,

Rajasekar