System and method for personalized search, information filtering, and for generating recommendations utilizing statistical latent class models

• US 7,328,216 B2
• Filed: 08/11/2003
• Issued: 02/05/2008
• Est. Priority Date: 07/26/2000
• Status: Expired due to Term

• Current Assignee: Open Text Holdings, Inc. (Open Text Corporation)
• Original Assignee: Recommind Incorporated (Open Text Corporation)
• Inventors: Hofmann, Thomas, Puzicha, Jan Christian
• Primary Examiner(s): Abel-Jalil; Neveen
• Application Number:
  US10/639,024
• Publication Number:
  US 20040034652A1
• Time in Patent Office: 1,639 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06F 16/335 Filtering based on additional data, e.g. user or group profiles filtering in web context G06F16/9535, G06F16/9536
• CPC Class Codes: Y10S 707/99934 Query formulation, input preparation, or translation
• CPC Class Codes: Y10S 707/99936 Pattern matching access
• CPC Class Codes: Y10S 707/99942 Manipulating data structure, e.g. compression, compaction, compilation

Abstract

The system implements a novel method for personalized filtering of information and automated generation of user-specific recommendations. The system uses a statistical latent class model, also known as Probabilistic Latent Semantic Analysis, to integrate data including textual and other content descriptions of items to be searched, user profiles, demographic information, query logs of previous searches, and explicit user ratings of items. The system learns one or more statistical models based on available data. The learning may be reiterated once additional data is available. The statistical model, once learned, is utilized in various ways: to make predictions about item relevance and user preferences on un-rated items, to generate recommendation lists of items, to generate personalized search result lists, to disambiguate a users query, to refine a search, to compute similarities between items or users, and for data mining purposes such as identifying user communities.

Claims

1. A method in a computer system for generating a recommendation list of desired items from a set of data including at least one of:

items, content descriptors for the items, user profiles about transactions, prior searches, user ratings and user actions, to generate a recommendation list of desired items, comprising the steps of;

statistically analyzing the set of data to learn semantic associations between words within specific items of the set of data, storing the learned semantic associations in computer hardware;

computing probabilities of each learned semantic association;

receiving into the computer system for generating a recommendation list;

an actual user profile, a user query; and

a request to generate at least one recommendation list, items in the recommendation list being ranked by their likelihood of being the desired items;

computing a probability of relevance of each item in the set of data to the actual user profile and said user query, the step of computing a probability of relevance including combining the learned semantic associations and the actual user profile; and

generating at least one recommendation list of desired items.

2. A method in a computer system containing a recommendation system for a set of data including items, content descriptors for the items, user profiles about transactions, prior searches, user ratings or user actions, to generate a recommendation list of desired items, comprising the following steps:

receiving into the recommendation system a set of statistical latent class models along with appropriate model combination weights, each possible combination of items, content descriptors, users, object attributes or user attributes, and preferences being assigned a probability indicating the likelihood of that particular combination, storing the set of statistical latent class models in computer hardware;

receiving into the recommendation system at least one of;

an actual user profile; and

a user query;

receiving into the recommendation system a request to generate at least one recommendation list, items in the recommendation list being ranked by their likelihood of being desired items

when said user query is not received into the recommendation system, generating a user query;

computing a probability of relevance for each item in the set of data utilizing the received set of statistical latent class models and the set of data, the step of computing a probability of relevance includes accounting for a temporal structure of transactions, ratings and user actions; and

computing the probabilities of relevance of items to a user for a particular point in time; and

generating at least one recommendation list, each recommendation list having a variable length and consisting of a ranked list of desired items, the items being ranking based on the computed probability of relevance.

3. A method in a computer system containing a recommendation system for a set of data including items, content descriptors for the items, user profiles about transactions, prior searches, user ratings or user actions, to generate a recommendation list of desired items, comprising the following steps:

receiving into the recommendation system a set of statistical latent class models along with appropriate model combination weights, each possible combination of items, content descriptors, users, object attributes or user attributes, and preferences being assigned a probability indicating the likelihood of that particular combination, storing the set of statistical latent class models in computer hardware;

receiving into the recommendation system at least one of;

an actual user profile; and

a user query;

receiving into the recommendation system a request to generate at least one recommendation list, items in the recommendation list being ranked by their likelihood of being desired items;

when said user query is not received into the recommendation system, generating a user query;

computing a probability of relevance for each item in the set of data utilizing the received set of statistical latent class models and the set of data, the step of computing a probability of relevancy includes computing expected utility to the user using the probability of relevance for each of the items; and

generating at least one recommendation list, each recommendation list having a variable length and consisting of a ranked list of desired items, the items being ranking based on the computed probability of relevance.

4. The method according to claim 3, wherein the at least one returned recommendation list further includes a list of object attributes, the attributes being determined by the following steps:

for each recommendation list, computing for each object attribute, a probability of the object attribute occurring in one of the items in the recommendation list based on the received set of statistical latent class models, weighted with their respective probabilities of relevance, and weighted with prior object attribute weights;

sorting the object attributes in each recommendation list; and

returning at least a higher probability portion of a sorted list of object attributes with the recommendation list.

5. The method according to claim 3, further comprising:

recognizing that the user query is ambiguous based on the received set of statistical latent class models; and

when the user query is ambiguous, distributing the items over multiple recommendation lists.

6. The method according to claim 3, further comprising:

computing predictions of unobserved variables.

7. A method in a computer system containing a recommendation system for a set of data including items, content descriptors for the items, user profiles about transactions, prior searches, user ratings or user actions, to generate a recommendation list of desired items, comprising the following steps:

generating a set of statistical latent class models by probabilistic latent semantic indexing and storing the set of statistical latent class models in computer hardware;

receiving into the recommendation system the set of statistical latent class models along with appropriate model combination weights, each possible combination of items, content descriptors, users, object attributes and user attributes, and preferences being assigned a probability indicating the likelihood of that particular combination;

receiving into the recommendation system at least one of;

an actual user profile; and

a user query;

receiving into the recommendation system a request to generate at least one recommendation list, items in the recommendation list being ranked by their likelihood of being desired items;

when said user query is not received into the recommendation system, generating a user query;

computing a probability of relevance for each item in the set of data utilizing the received set of statistical latent class models and the set of data; and

generating at least one recommendation list, each recommendation list having a variable length and consisting of a ranked list of desired items, the items being ranking based on the computed probability of relevance.

8. The method according to claim 7, wherein the step of receiving into the recommendation system at least one of:

an actual user profile; and

a user query comprises receiving an actual user profile, the received actual user profile including a prior history.

9. The method according to claim 7, wherein the at least one returned recommendation list further includes a list of object attributes, the object attributes being determined by the following steps:

for each recommendation list, computing for each object attribute, a probability of the object attribute occupying in one of the items in the recommendation list based on the received set of statistical latent class models, weighted with their respective probabilities of relevance, and weighted with prior object attribute weights;

sorting the object attributes in each recommendation list; and

returning at least a higher probability portion of a sorted list of object attributes with the recommendation list.

10. The method according to claim 9, wherein the object attributes include keywords.

11. The method according to claim 7, further comprising:

recognizing that the user query is ambiguous based on the received set of statistical latent class models; and

when the user query is ambiguous, distributing the items over multiple recommendation lists.

12. The method according to claim 7, further comprising:

computing predictions of unobserved variables.

13. The method according to claim 12, wherein the set of data includes user ratings and the unobserved variables include at least one of user ratings on items and query terms.

14. The method according to claim 7, wherein prior to the step of receiving a set of statistical latent class models, learning user model parameters.

15. The method according to claim 7, wherein the set of received models form a hierarchy that is used in successive refinement of the user query.

16. A method in a computer system containing a recommendation system for a set of data including items, content descriptors for the items, user profiles about transactions, prior searches, user ratings or user actions, to generate a recommendation list of desired items, comprising the following steps:

receiving into the recommendation system a set of statistical latent class models along with appropriate model combination weights, each possible combination of items, content descriptors, users, object attributes and user attributes, and preferences being assigned a probability indicating the likelihood of that particular combination, storing the set of statistical latent class models in computer hardware;

receiving into the recommendation system at least one of;

an actual user profile; and

a user query;

receiving into the recommendation system a request to generate at least one recommendation list, items in the recommendation list being ranked by their likelihood of being desired items;

when said user query is not received into the recommendation system, generating a user query;

prior to the step of computing a probability of relevance, setting preference probabilities below a predetermined threshold to zero,

computing a probability of relevance for each item in the set of data utilizing the received set of statistical latent class models and the set of data; and

generating at least one recommendation list, each recommendation list having a variable length and consisting of a ranked list of desired items, the items being ranking based on the computed probability of relevance.

17. The method according to claim 16, wherein the at least one returned recommendation list further includes a list of object attributes, the object attributes being determined by the following steps:

for each recommendation list, computing for each object attribute, a probability of the object attribute occupying in one of the items in the recommendation list based on the received set of statistical latent class models, weighted with their respective probabilities of relevance, and weighted with prior object attribute weights;

sorting the object attributes in each recommendation list; and

returning at least a higher probability portion of a sorted list of object attributes with the recommendation list.

18. The method according to claim 16, further comprising:

recognizing that the user query is ambiguous based on the received set of statistical latent class models; and

when the user query is ambiguous, distributing the items over multiple recommendation lists.

19. The method according to claim 16, further comprising:

computing predictions of unobserved variables.