Data processing: financial – business practice – management – or co – Automated electrical financial or business practice or... – Electronic shopping
Reexamination Certificate
1999-09-03
2003-10-21
Coggins, Wynn W. (Department: 3625)
Data processing: financial, business practice, management, or co
Automated electrical financial or business practice or...
Electronic shopping
C706S012000, C706S014000, C707S793000, C707S793000
Reexamination Certificate
active
06636836
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a recommendation system as one of information filtering methods and more particularly to a system for recommending items with multiple analyzing components, for allowing a combination of various recommendation methods to be used by employing multiple recommendation agents and adjusting a value of influential power of each recommendation agent on a user to reflect the adjusted value on a subsequent recommendation.
2. Description of Related Art “Information overload” is not a new story for the moderns any more and the amount of information is rapidly increasing. Automatic “information filtering” is one of the efforts to obtain appropriate information from the excessive information.
In a wide meaning, the information filtering denotes every action for sorting out only necessary information for a user from overflowing information. Information filtering techniques are widely adopted into range from electronic mail or news at an early stage to web files and articles on electronic commerce these days, reflecting current information resources and aspects of information consumption. In particular, as the electronic commerce and internet acceleratedly grow up lately, the information filtering techniques are more acutely required than has ever been.
There are a content-based filtering method and a collaborative filtering method in conventional information filtering methods.
The content-based filtering method has been most widely used since it made its first appearance. Briefly describing the content-based filtering method, characteristics of contents of items to be estimated are programmed and discerned and a value of each item is determined based upon characteristics favored by a user. This method has obtained great results in searching files on the web or filtering e-mail.
However, this content-based filtering method can be applied only to a case when contents are recognizable through a computer, e.g., texts. There is a trouble in applying the content-based filtering method to items whose contents are difficult to analyze, such as movies, music and food. Moreover, this method recommends only items that a user has already used or that are similar to those used by the user. Accordingly, novel or accidental discovery is difficult to occur, and user feedback for constituting and developing the user's profile should be obtained.
The collaborative filtering is provided to supplement disadvantages of the content-based filtering. The collaborative filtering method is based upon an idea that a person can make the better decision than the computer although it is substantially slow. In the collaborative filtering method, decision of values (estimation) of items to be filtered is not assigned to a program but to the person and the result of the estimation is used in the program. In other words, the person estimates items independently and the computer automatically collects the estimations and finds persons who have similar tastes. The result of the operations by the computer is used for making a decision related to a particular person or judging a value of an item. The collaborative filtering method may be applied to filter usual texts as well as fields in which the content-based filtering is weak, such as music, movies, restaurants and places recommendable for travel. The collaborative filtering method is disclosed in U.S. Pat. Nos. 4,870,579 and 4,996,642.
However, the collaborative filtering method also needs to obtain user feedback for constituting and developing user profiles. Furthermore, if a new user or item appears, recommendation or estimation becomes difficult. This makes it difficult to appropriately recommend items to a user having a peculiar taste. Because information on characteristics of items is not considered, the probability of giving a better recommendation is abandoned.
Recently, there have been attempts to combine both the two methods. However, most of attempts just partially use one technique based upon the other technique or are restricted a lot in applied fields. Such study is just for improving application techniques themselves but is lack in considering users who consume recommendations. As a result, the recent study is not enough to overcome a structural limit that a particular application has.
SUMMARY OF THE INVENTION
Accordingly, the present invention is directed to a system for recommending items with multiple analyzing components, which substantially obviates one or more of the limitations and disadvantages of the related art.
An objective of the present invention is to provide a system for individually serving users with electronic contents including electronic commerce and internet services using an automatic filtering technique and for recommending interesting items to each user, such as goods and service contents in an intellectual manner.
Another objective of the present invention is to provide a recommendation system for integrally applying a variety of analyzing methods to items and for allowing each analyzing method to be personalized for a particular user and to be discriminately applied to each user.
Additional features and advantages of the invention will be set forth in the following description, and in part will be apparent from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure as illustrated in the written description and claims hereof, as well as the appended drawings.
To achieve these and other advantages, and in accordance with the purpose of the present invention as embodied and broadly described, a system for recommending items with multiple recommendation agents comprises: the multiple recommendation agents each for providing a list of recommended items for each user and auxiliary information of the recommended items in its independent manner; a user agent for holding basic information on each user and influential power information of each recommendation agent with respect to each user; and a recommendation manager for selecting one or more recommendation agents from the plurality of recommendation agents using the basic user information and the influential power information of each recommendation agent with respect to each user received from the user agent, generates a final list of recommended items from the list or lists of recommended items suggested by the selected recommendation agent or agents using the basic user information and the influential power of each recommendation agent received from the user agent and the auxiliary information received from the selected recommendation agent or agents, and adjusting the influential power of each recommendation agent stored in the user agent according to a result of the recommendation.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory and are intended to provide further explanation of the invention as claimed.
REFERENCES:
patent: 4870579 (1989-09-01), Hey
patent: 4996642 (1991-02-01), Hey
patent: 5634051 (1997-05-01), Thomson
patent: 5659732 (1997-08-01), Kirsch
patent: 5864845 (1999-01-01), Vorhees et al.
patent: 5873080 (1999-02-01), Coden et al.
patent: 5905862 (1999-05-01), Hoekstra
patent: 5987446 (1999-11-01), Corey et al.
patent: 6006218 (1999-12-01), Breese et al.
patent: 6041311 (2000-03-01), Chislenko et al.
patent: 6064980 (2000-05-01), Jacobi et al.
patent: 6078914 (2000-06-01), Redfern
patent: 6192354 (2001-02-01), Bigus et al.
patent: 6266649 (2001-07-01), Linden et al.
patent: 6266668 (2001-07-01), Vanderveldt et al.
patent: 6397212 (2002-05-01), Biffar
patent: 822502 (1998-04-01), None
Rudich, J. Shopbots. (intelligent robotic agents), Link-Up vol. 16, No. 4 (Jul.-Aug. 1999) p. 26.*
Torode, C. Intelligent agents' for real?, Computer Reseller News, No. 706 (Oct. 21, 1996) p. 63.*
Wisewire to Present at Internet Showcase: Launches Personalized Internet Information Service Based on
Brown Tim
Coggins Wynn W.
Iwingz Co., Ltd.
Perman & Green LLP
LandOfFree
Computer readable medium for recommending items with... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Computer readable medium for recommending items with..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Computer readable medium for recommending items with... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3120737