skip to navigation
skip to content

Not Logged In

cold-start-recommender 0.3.15

In-memory recommender for recommendations produced on-the-fly

“Will it scale?” is a less important question than “will it ever matter?” [kadavy.net]

We developed Cold Start Recommender because we needed a recommender with the following characteristics:

  • Greedy. Useful in situations where no previous data on Items or

    Users are available, therefore any information must be used –not just which Item a User likes, but also –in the case of a book– the corresponding category, author etc.

  • Fast. Any information on Users and Item should be stored and

    used immediately. A rating by any User should improve recommendations for this User, but also for other Users. This means in-memory database and no batch computations.

  • Ready to use. Take a look at recommender_api.py to start

    a webapp that POSTs information and GETs recommendations.

CSRec should not (yet) be used for production systems, but only for pilots, where statistics are so low that filters (e.g. loglikelihood filter on the co-occurence matrix) are premature. It aims to gather data in order to immediately personalise the user experience.

TODO Future releases will include state of the art algorithms (availability planned for October 2014, a few months before our product goes public).

CSRec is written in Python, and under the hood it uses the Pandas library.

Table of Contents

A simple script

from csrec.Recommender import Recommender

engine = Recommender()

# Insert Item with it properties (e.g. author, category…) # NB lists can be passed as json-parseable strings engine.insert_item({‘_id’: ‘an_item’, ‘author’: ‘The Author’, ‘tags’: ‘[“nice”, “good”]’})

# Insert rating, indicating wich property of the Item should be used for producing recs

engine.insert_rating(user_id=’a_user’, item_id=’an_item’, rating=4, item_info=[‘author’, ‘tags’])

# Insert rating, indicating that only the property should be used for recs (e.g. initial users’ profiling)

engine.insert_rating(user_id=’another_user’, item_id=’an_item’, rating=3, item_info=[‘author’], only_info=True)

Features

Persistence

You can use CSRec purely in-memory for testing or with MongoDB, which you can install on a tmpfs filesystem created in your RAM (on Linux, see http://edgystuff.tumblr.com/post/49304254688/how-to-use-mongodb-as-a-pure-in-memory-db-redis-style). If using a RAM partition, please make a replica set!

(Why use a replica set? Because you can have the primary DB in memory, and two other secondaries on disk. If the primary goes down, you still can use CSRec at lower performances, but without any data loss.)

Examples:

engine = Recommender() # Start in-memory recommender for testing

engine = Recommender(mongo_host=’localhost’, mongo_db_name=’my_cold_rec’) # …with MongoDB, collections are created automatically

engine = Recommender(mongo_host=’localhost’, mongo_db_name=’my_cold_rec’, mongo_replica_set=’recommender_replica’) # as above, with replica

The Cold Start Problem

The Cold Start Problem originates from the fact that collaborative filtering recommenders need data to build recommendations. Typically, if Users who liked item ‘A’ also liked item ‘B’, the recommender would recommend ‘B’ to a user who just liked ‘A’. But if you have no previous rating by any User, you cannot make any recommendations.

CSRec tackles the issue in various ways:

1. It allows profiling with well-known Items without biasing the results. For instance, if a call to insert_rating is done in this way:

engine.insert_rating(user_id=’another_user’, item_id=’an_item’, rating=3, item_info=[‘author’], only_info=True)

CSRec will only register that ‘another_user’ likes a certain author, but not that s/he might like ‘an_item’. This is of fundamental importance when profiling Users with a “profiling page” on your website. If you ask Users whether they prefer “Harry Potter” or “The Better Angels of Our Nature”, and most of them choose Harry Potter, you would not want to make the Item “Harry Potter” even more popular. You might just want to record that those users like children’s books marketed as adult literature.

CSRec does that because, unless you are Amazon or a similar brand, the co-occurence matrix is often too sparse to compute decent recommendations. In this way you start building multiple, denser, co-occurence matrices and use them from the very beginning.

2. Any information is used. You decide which information you should record about a User rating an Item. This is similar to the previous point, but you also register the item_id.

3. Any information is used *immediately*. The co-occurence matrix is updated as soon as a rating is inserted.

4. It tracks anonymous users, e.g. random visitors of a website before the sign in/ sign up process. After sign up/ sign in the information can be reconciled –information relative to the session ID is moved into the correspondent user ID entry.

Algorithms

At the moment CSRec only provides purely item-based recommendations (co-occurence matrix dot the User’s ratings array). In this way we can provide recommendations in less than 200msec for a matrix of about 10,000 items.

Versions

v 3.14

Fixed logging and bug for computing cooccurrence matrix

v 3.13

Added self.drop_db

v 3.12

  • Bug fixed

v 3.11

  • Some debugs messsages added

v 3.10

v 3.8

  • Sync categories’ users and items collections in get_recommendations

v 3.7

Bug fixing for in-memory

v 3.5

  • Added logging
  • Added creation of collections for super-cold start (not even one rating, and still user asking for recommendations…)
  • Additional info used for recommendations (eg Authors etc) are now stored in the DB
  • _sync_user_item_ratings now syncs addition info’s collections too
  • popular_items now are always returned, even in case of no rating done, and get_recommendations eventually adjusts the order if some profiling has been done
 
File Type Py Version Uploaded on Size
cold-start-recommender-0.3.15.tar.gz (md5) Source 2015-01-19 11KB
  • Downloads (All Versions):
  • 86 downloads in the last day
  • 351 downloads in the last week
  • 1133 downloads in the last month