Effective library recommending a product based on user history

Question

I have a database of which products are viewed by each user, and I want to recommend a product based on what similar users viewed. Is there a Python library that can do this? I don't need Netflix quality results, just products that are more than interesting. Any ideas?

Answer 1

You can check out pysuggest .

From the site:

SUGGEST is a Top-N recommendation engine that implements many recommendation algorithms. Top-N recommendation systems, personalized information filtering technologies, are used to identify a set of N elements that will be of interest to a specific user. In recent years, the Supreme Advisor systems have been used in a number of different applications, such as recommending products that the client will most likely buy; we recommend films, TV programs or music that the user will find pleasant; identify web pages that will be of interest; or even suggest alternative ways of finding information.

Answer 2

k-Nearest Neighbor is probably the most frequently used algorithm for real-time online recommendations.

In NumPy / SciPy you have several options [note: the answer was updated on December 12 to reflect updates in the sklearn library]:

• nearest neighbors module in scikit-learn (aka sklearn ); This is a complex implementation of kNN, which includes the offset and voting of a neighbor, as well as a very efficient storage / retrieval component (spherical tree);

• scipy.spatial . I used this for several projects, although it is unlikely that I will use it for my next, which is now available for sklearn, which is a more modern implementation of kNN. However, the spatial module has the kd-tree class (instead of the standard numpy array, kd-tree is used by this module to store data, which along w / Voronoi teselation is the most common specialized data structure for storing very large data sets for kNN); in addition, it has methods for several distance metrics (other than Euclidean distance).

Answer 3

Here's another python library for implementing a recommendation system:

ocelma / python-recsys

This is a very simple use! Examples here: Quick Start¶