Automated decision advisor

• US 7,080,071 B2
• Filed: 08/06/2001
• Issued: 07/18/2006
• Est. Priority Date: 08/04/2000
• Status: Expired due to Term

• Current Assignee: IAC Search & Media Incorporated (Match Group, Inc.)
• Original Assignee: Ask Jeeves, Inc. (IAC, Inc.)
• Inventors: Robinson, Zac, Chrisman, Lonnie, Henrion, Max
• Primary Examiner(s): Rones, Charles
• Assistant Examiner(s): ABEL JALIL, NEVEEN
• Application Number:
  US09/923,772
• Publication Number:
  US 20020077931A1
• Time in Patent Office: 1,807 Days
• Field of Search: 725/116, 707/3, 707/8
• US Class Current: 707/713
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: Y10S 707/99932 Access augmentation or optimizing
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 707/99936 Pattern matching access

Abstract

The present invention provides an automated system to guide people (“users”) to help them find and select products, services, or other items, that best meet their needs and preferences. The system obtains information from and about users to assess their needs and preferences. The system dynamically selects those questions to ask that are most likely to help discriminate between items based on information about user preferences and differences between the items, based on information about the user obtained so far. The system scores the available items in terms of how well they match user'"'"'s needs and preferences and generates lists of recommended items. The system explains its recommendations using lists of pros and cons to help users understand how well the features or attributes of the items do or do not match their expressed needs.

Claims

1. A computer-implemented method for assisting a user in determining one or more items from an item set, the method comprising:

outputting one or more questions about a user'"'"'s preferences;

receiving answers for the one or more questions related to the user'"'"'s preferences;

for each answer to a question in the one or more questions, performing the steps of;

determining one or more attributes that are associated with the question; and

updating or specifying a relative importance value for the one or more attributes based on the answer to the question, wherein a relative importance value is updated for an attribute when the attribute was associated with a prior question and specified when the attribute was not associated with a prior question;

generating at least one list of items selected from the item set based on the relative importance values for the one or more attributes determined for the answers for the one or more questions; and

outputting the at least one list of items.

2. The method of claim 1, further comprising determining at least one pro and at least one con for an attribute, the at least one pro and at least one con identifying how well the attribute meets the user'"'"'s preferences.

3. The method of claim 2, further comprising generating a statement that includes a recommendation based on the determined at least one pro and the at least one con.

4. The method of claim 1, wherein outputting the one or more questions comprises outputting a series of questions, wherein questions are dynamically determined based on prior answers to prior questions.

5. The method of claim 4, wherein the dynamically determined questions are determined based on their likely affect on the relative importance values for the one or more attributes.

7. The method of claim 1, wherein the item is a product, service or other selectable element.

8. The method of claim 1, wherein the information obtained from questions to the user is supplemented with other information about the user.

9. The method of claim 8, wherein the other information includes one or more of information provided by the user during online interactions and information provided by the user during offline interactions.

10. The method of claim 9, wherein the information provided by the user is one or more of registration information, information about prior selections, information about prior purchases and information about prior interactions with automated decision advisors.

11. The method of claim 1, further comprising a step of providing the user with explanations of recommendations, where each explanation contains at least one pro or at least one con, and the explanation refers to an attribute of the item and identifies how well the attribute meets the user'"'"'s set of target preferences.

12. The method of claim 11, wherein one or more pro or con includes a reference to the set of target preferences.

13. The method of claim 11, further comprising generating a statement relating one or more pro or con to a matching of an attribute and a recommendation for or against an item based on indirect inferences from user inputs.

14. The method of claim 13, wherein each explanation includes those pros and cons that have high estimated decision relevance to the user, and excludes at least one mention of a low estimated relevance attribute.

15. The method of claim 1, further comprising generating for display to the user a list of common attributes for which all of the items on the list of items have common values.

16. The method of claim 15, wherein the list of common attributes includes attributes that have values in a common range.

17. The method of claim 15, wherein the list of common attributes includes pros and cons for attributes that have high estimated importance to the user and excludes at least one pro or con for at least one attribute that has a low estimated importance to the user.

18. The method of claim 17, wherein one or more pro or con includes a reference to the set of target preferences.

19. The method of claim 15, further comprising generating a statement relating one or more pro or con to a matching of an attribute and a recommendation for or against an item based on indirect inferences from user inputs.

20. The method of claim 1, further comprising offering questions to the user in a sequence that is determined dynamically based on answers to previous questions.

21. The method of claim 20, wherein sequence of questions, is based on a priority calculated as an increasing function of the uncertainty about the importance of the attributes to which the question pertains and the variation in attribute relative importance values among the items.

22. The method of claim 21, wherein the variation of attribute relative importance value among the items is weighted to the values of the items with the high current ranking so that the question priority reflects the expectation that answers to the question are likely to change the ranking of the top ranked items.

23. The method of claim 22, wherein the system mentions for each unmet requirements and selected combinations or requirements the number of items it or they eliminate as unacceptable.

24. The method of claim 1, wherein if no item meets all requirements specified by the user, the system displays a conflict page that mentions the fact that no items meet all requirements and lists the requirements that cannot be simultaneously met.

25. The method of claim 1, wherein if no item meets all requirements specified by the user, the system determines a best match based on the one or more relative importance values for the one or more attributes.

6. A computer-implemented method for assisting a user in selecting one or more items from an item set that best match a set of target preferences, the method comprising:

obtaining information from the user about that user'"'"'s set of target preferences, using direct questions;

associating the set of target preferences with one or more attributes;

determining one or more relative importance values for the one or more attributes based on the obtained information;

estimating decision relevance of an attribute to the user as an increasing function of estimated importance of the attribute to the user and an absolute relative preference value, wherein the absolute relative preference value of a given item is a measure of a difference between an attribute value for the given item and an average value of the attribute for items in the list of items other than the given item;

generating at least one list of items selected from the item set that will best meet the set of target preferences based on the one or more relative importance values for the one or more attributes; and

generating a display including the list of items and explanations for at least one item on the list of items explaining how well one or more attributes of the at least one item match the set of target preferences.