Expert supported interactive product selection and recommendation

• US 7,558,773 B2
• Filed: 05/10/2007
• Issued: 07/07/2009
• Est. Priority Date: 07/19/2000
• Status: Active Grant

• Current Assignee: NetCracker Technology Corporation (NEC Corporation)
• Original Assignee: Convergys Corporation
• Inventors: Mancisidor, Rod, Erickson, Charles R., Gheith, Ahmed, Chan, William W.
• Primary Examiner(s): STARKS, WILBERT L
• Application Number:
  US11/746,767
• Publication Number:
  US 20070208682A1
• Time in Patent Office: 789 Days
• Field of Search: 706/52, 706/45
• US Class Current: 706/52
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: G06Q 10/10 Office automation; Time management
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

Expert System supported interactive product selection and recommendation. The invention assists an agent to interact with a customer and to provide selection and recommendation of available products and/or services that offer a workable solution for the customer. The invention allows for the use of agents of varying skill levels, including relatively low skill level, without suffering deleterious performance. From certain perspectives, an expert system employed using various aspects of the invention allows the agent to provide real time interaction with a customer and to provide a real time recommended solution to that customer. Many traditional approaches dealing in complex industries require that agent'"'"'s have a high degree of skill and expertise. The invention allows even a novice agent to service a customer'"'"'s needs without requiring a high skill level or up-front training that is often at the expense of the provider seeking to market its products and/or services.

Claims

1. A system for rating a product based on a need of a user and a trait of the product, the system comprising:

a) means for obtaining a plurality of values using a plurality of set operations performed on a first fuzzy set representing the need of the user and a second fuzzy set representing the trait of the product;

b) a computer readable medium having stored thereon an equation, said equation comprising at least two variables corresponding to values from said plurality of values;

c) a processor configured to rate said product with respect to the need of the user represented by the first fuzzy set based at least in part on the equation stored on the computer readable medium.

2. The system as claimed in claim 1 wherein the product comprises a service.

3. The system as claimed in claim 2 wherein using the plurality of set operations comprises:

a) performing a set operation a first time using a first input and

b) performing the set operation a second time using a second input.

4. The system as claimed in claim 2 wherein the function of obtaining the plurality of values using the plurality of set operations performed on the first fuzzy set and the second fuzzy set comprises performing the following operations:

a) (P ∩

N)and

b) ( P ∩

N)

wherein N represents the first fuzzy set, and P represents the second fuzzy set.

5. The system as claimed in claim 2 wherein the equation comprises dividing a first variable representing a likelihood that the user will be satisfied with the trait of the product with a combination of the first variable and a second variable representing a likelihood that the user will not be satisfied with the trait of the product.

6. The system as claimed in claim 2 wherein the first fuzzy set representing the need of the user is a triangular fuzzy set.

7. The system as claimed in claim 2 wherein the second fuzzy set representing the trait of the product is chosen from the group consisting of an S fuzzy set and a Z fuzzy set.

8. The system as claimed in claim 2 wherein the trait of the product is a domain specific trait.

9. A machine for rating a product based on a need of a user and a trait of the product, the machine comprising:

a) a processor;

b) a memory; and

c) a display;

wherein;

a) said memory stores a set of data comprising, for each of a plurality of products, a plurality of fuzzy sets, each of said fuzzy sets representing a level of satisfaction with respect to a characteristic of relevance for the product;

b) said processor is configured with a set of instructions operable to configure the processor to perform a set of functions comprising;

i) generating a rating for a product from the plurality of products with respect to a specific characteristic of relevance by using the fuzzy set stored in the memory representing the level of satisfaction with respect to the specific characteristic of relevance for the product, a fuzzy set representing the need of the user with respect to the specific characteristic of relevance, and a fuzzy set representing a level of dissatisfaction with respect to the specific characteristic of relevance for the product; and

ii) sending a set of display instructions to the display configured to cause the display to display a recommendation output, said recommendation output being based at least in part on said rating, and comprising a recommendation of at least one product from the plurality of products;

c) said display is communicatively connected to said processor, and is operable to display said recommendation output in accordance with said set of display instructions.

10. The machine of claim 9, wherein generating a rating for a product from the plurality of products with respect to a specific characteristic by using the fuzzy set from the plurality of fuzzy sets stored in the memory representing the level of satisfaction with respect to the specific characteristic of relevance for the product, a fuzzy set representing the need of the user with respect to the specific characteristic of relevance, and a fuzzy set representing a level of dissatisfaction with respect to the specific characteristic of relevance for the product comprises:

a) determining a good value by performing a first set of one or more calculations comprising calculating h(P ∩

N);

b) determining a bad value by performing a second set of one or more calculations comprising calculating h(D ∩

N);

c) dividing the good value by the sum of the good value and the bad value;

wherein;

a) N represents the fuzzy set representing the need of the user with respect to the specific characteristic of relevance for the product;

b) P represents the fuzzy set from the plurality of fuzzy sets stored in the memory representing the level of satisfaction with respect to the specific characteristic of relevance for the product;

c) D represents the fuzzy set representing the level of dissatisfaction with respect to the specific characteristic of relevance for the product.

11. The machine of claim 10, wherein the fuzzy set representing the level of dissatisfaction with respect to the specific characteristic of relevance for the product is the compliment of the fuzzy set from the plurality of fuzzy sets stored in memory representing the level of satisfaction with respect to the specific characteristic of relevance for the product.

12. The machine of claim 10, wherein the fuzzy set representing the level of dissatisfaction with respect to the specific characteristic of relevance for the product is retrieved from said memory, and wherein said memory is nonvolatile.

13. The machine of claim 10, wherein the first set of one or more calculations consists of calculating h(P ∩

N) and wherein the second set of one or more calculations consists of calculating h(D ∩

N).

14. A computer readable medium having stored thereon a set of computer executable instructions operable to configure a computer to rate a product based on a need of a user and a trait of the product by performing a set of functions comprising:

a) generating a rating for the product with respect to a specific characteristic of relevance by using a fuzzy set representing a level of satisfaction with respect to the specific characteristic of relevance for the product, a fuzzy set representing the need of the user with respect to the specific characteristic of relevance, and a fuzzy set representing a level of dissatisfaction with respect to the specific characteristic of relevance for the product; and

b) sending a set of display instructions to a display configured to cause the display to display a recommendation output, said recommendation output being based at least in part on said rating.

15. The computer readable medium of claim 14, wherein generating a rating for the product with respect to a specific characteristic by using a fuzzy representing the level of satisfaction with respect to the specific characteristic of relevance for the product, a fuzzy set representing the need of the user with respect to the specific characteristic of relevance, and a fuzzy set representing a level of dissatisfaction with respect to the specific characteristic of relevance for the product comprises:

a) determining a good value by performing a first set of one or more calculations comprising calculating h(P ∩

N);

b) determining a bad value by performing a second set of one or more calculations comprising calculating h(D ∩

N);

c) dividing the good value by the sum of the good value and the bad value;

wherein;

a) N represents the fuzzy set representing the need of the user with respect to the specific characteristic of relevance for the product;

b) P represents the fuzzy set representing the level of satisfaction with respect to the specific characteristic of relevance for the product;

c) D represents the fuzzy set representing the level of dissatisfaction with respect to the specific characteristic of relevance for the product.

16. The computer readable medium of claim 15 wherein the set of functions comprises obtaining the fuzzy set representing the level of dissatisfaction with respect to the specific characteristic of relevance for the product by calculating the compliment of the fuzzy set representing the level of satisfaction with respect to the specific characteristic of relevance for the product.

17. The computer readable medium of claim 15, wherein the first set of one or more calculations consists of calculating h(P ∩

N) and wherein the second set of one or more calculations consists of calculating h(D ∩

N).

18. The computer readable medium of claim 15, wherein the set of functions comprises retrieving the fuzzy set representing the level of satisfaction with respect to the specific characteristic of relevance for the product from a nonvolatile memory.

19. The computer readable medium of claim 18, wherein the set of functions comprises retrieving the fuzzy set representing the level of dissatisfaction with respect to the specific characteristic of relevance for the product is retrieved from said nonvolatile memory.