Method and Apparatus for Executing a Recommendation

• US 20120185481A1
• Filed: 09/21/2009
• Published: 07/19/2012
• Est. Priority Date: 09/21/2009
• Status: Abandoned Application

• Current Assignee: Telefonaktiebolaget LM Ericsson
• Original Assignee: Telefonaktiebolaget LM Ericsson
• Inventors: Björk, Jonas, Lidström, Mattias, Söderberg, Joakim, Moritz, Simon
• Application Number:
  US13/496,900
• Publication Number:
  US 20120185481A1
• Time in Patent Office: Days
• US Class Current: 707/737
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

A method, apparatus and system for generating recommendations of items to users. Ratings of items made by users are collected (1:1). User behaviour information is also collected (1:2). Then correlations in ratings are obtained (1:3) and similarities in user behaviour amongst the users are obtained (1:4). Thereafter, an item for recommendation to a user is identified (1:5), based on both the correlations in ratings and on the similarities in user behaviour amongst the users and the item is recommended to the user (1:6).

Claims

1-12. -12. (canceled)

13. An apparatus configured to identify items for recommendation to a user and recommend said items to said user, said apparatus comprising:

a collecting unit configured to collect ratings of items made by users and to collect user behavior information;

an obtaining unit configured to obtain correlations in ratings by computing the correlations and to obtain similarities in user behavior amongst the users by computing the similarities, by clustering similar users together using machine learning techniques;

an identifying unit configured to identify an item for recommendation to a user, based on both the computed correlations in ratings and on the computed similarities in user behavior; and

a recommending unit configured to recommend said item to said user.

14. The apparatus of claim 13, wherein the apparatus is further configured to compute the similarities in user behavior amongst the users by clustering similar users together using one or more machine learning techniques, including at least one of:

K-means clustering methods, support vector machine methods, Latent Semantic Analysis (LSA), and Probabilistic Latent Semantic Analysis (PLSA).

15. The apparatus of claim 13, wherein the apparatus is further configured to retrieve feedback from users, the feedback relating to previously recommended items.

16. The apparatus of claim 15, wherein the apparatus is further configured to determine exploit and explore factors depending on the feedback and on the number of ratings performed by a user, wherein the exploit factor is related to a correlation in ratings and the explore factor is related to similarities in user behavior, in order to identify an item for recommendation to the user.

17. The apparatus of claim 16, wherein the apparatus is further configured to give more weight to the explore factor when a positive feedback, indicating that the user has consumed a previously recommended item, is received and to give less weight to the explore factor when a negative feedback, indicating that the user has not consumed a previously recommended item, is received.

18. The apparatus of claim 16, wherein the apparatus is further configured to adjust weights as a function of the exploit and explore factors, wherein the exploit factor is given more weight the more ratings the user has given and the explore factor is given more weight the less ratings the user has given.

19. The apparatus of claim 18, wherein the apparatus is further configured to predict items, for recommendation to the user, with the adjusted weights.

20. The apparatus of claim 19, wherein the apparatus is further configured to predict items for recommendation to the user using a nearest neighborhood algorithm.

21. The apparatus of claim 19, wherein the apparatus is further configured to produce recommendations by ranking the predicted values.

22. The apparatus of claim 13, wherein the apparatus is further configured to collect the user behavior information from at least one of:

Charging Data Records, Dynamic User Data Records, and Location Data Records.

23. A system configured to find an item or items for recommendation to a user, said system comprising:

a first database storing data, related to ratings of at least one of items and users;

a second database storing dynamic user data, related to user behavior information;

an apparatus configured to retrieve ratings of items or users from said first database and compute correlations in ratings;

an apparatus configured to retrieve user behavior information from said second database and to compute similarities in user behavior amongst the users by clustering similar users together using machine learning techniques; and

an apparatus configured to retrieve computed similarities in user behavior amongst users, to retrieve computed correlations in ratings, and to identify an item or items for recommendation to a user based on both the computed correlations in ratings and the computed similarities in user behavior.

24. The system of claim 23, further comprising Service Delivery Node for providing a service to the user and for requesting recommendations of items to the user.