Intelligent knowledge base content categorizer (IKBCC)

• US 6,782,391 B1
• Filed: 10/01/1999
• Issued: 08/24/2004
• Est. Priority Date: 10/01/1999
• Status: Expired due to Term

• Current Assignee: NCR Voyix Corporation
• Original Assignee: NCR Corporation
• Inventors: Scher, Karen M.
• Primary Examiner(s): KINDRED, ALFORD W
• Application Number:
  US09/410,994
• Time in Patent Office: 1,789 Days
• Field of Search: 707/1-7, 707/100-104, 707/205, 707/500, 707/524, 706/50, 709/200-203, 709/224, 715/500.1, 715/501.1, 715/526
• US Class Current: 707/603
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: Y10S 707/99943 Generating database or data structure, e.g. via user interface

Abstract

Disclosed is a method for improving the quality of Metadata and information resident on an Internet self-service resource such as a knowledge base (KB). The present invention assists in reducing and/or negating the time customer service agents spent to publish their responses to customer inquiries in a KB. An Intelligent Knowledge Base Content Categorizer (IKBCC) accomplishes both these goals by automatically, intelligently, and consistently evaluating the contents of customer service agent (CSA) responses to customer requests. It proactively determines the value of adding a response to the existing database, and methodically produces Metadata to describe each new KB entry. The IKBCC uses an algorithm that factors in variables (such as KB data, CSA information, administrative parameters, etc.), to evaluate whether a KB entry already exists or should be added. It then determines the nature of the Metadata to describe the added entries. This invention works as an integral part of an overall Internet self-service offering. And its method of indexing agent responses provides end users with a KB that allows them timelier, more relevant query feedback, by categorizing and organizing the data within an available Internet self-service resource.

Claims

1. A computer-implemented method of automatically adding data to a database using human agent responses to questions, comprising:

receiving a question and routing the question to a human agent;

using a response from said human agent to said question, determining whether to add the response to the database;

if said determining step is positive then creating metadata for categorizing and describing said response, and updating the database with the response and said metadata; and

if said determining step is negative then rejecting the response.

2. The method of claim 1, comprising parsing the response using at least one of keywords, key phrases, response categorizations, and natural language interpretation.

3. The method of claim 1, comprising comparing the response against previous agent responses to determine if the agent'"'"'s response is unique.

4. The method of claim 3, comprising parsing the response using at least one of keywords, key phrases, response categorizations, and natural language interpretation.

5. The method of claim 4, comprising comparing the results of the parsing of the agent'"'"'s response against previous results of the parsing the agent'"'"'s response.

6. The method of claim 5, comprising determining the percentage of exact corresponding and partially corresponding metadata and/or database entries.

7. The method of claim 1, comprising evaluating a credibility factor of the human agent providing the response.

8. The method of claim 7, wherein the credibility factor includes at least one of determining the agent'"'"'s area of expertise and determining the agent'"'"'s status.

9. The method of claim 1, comprising determining the frequency with which the agent'"'"'s previous responses have been stored in a knowledge base.

10. The method of claim 9, comprising assigning a period of time during which the frequency is determined.

11. The method of claim 1, comprising determining the frequency of recurring questions to the agent providing the response.

12. The method of claim 1, comprising determining the percentage of times percentage of times questions required escalation or reassignment from the current agent.

13. The method of claim 1, further comprising updating an entry associated with the human agent in an agent information database.

14. The method of claim 13, further comprising using the updated entry to determine credibility of the agent'"'"'s subsequent response to a subsequent question.

15. The method of claim 14, further comprising determining, based on the credibility of the agent'"'"'s subsequent response, whether to add the agent'"'"'s subsequent response to said database.

16. The method of claim 13, wherein said updating of the agent information database is performed regardless of whether said determining step is positive or negative.

17. The method of claim 13, wherein a credibility score of the agent stored in the agent information database is increased if said determining step is positive, and decreased if said determining step is negative.

18. The method of claim 1, further comprising using the database as a knowledge base for customer self-service purposes.

19. The method of claim 1, wherein said determining step is positive if the response does not have a corresponding entry in said database, and a credibility score of the agent satisfies a predetermined criterion.

20. The method of claim 1, wherein the response is quality checked and if passed is returned to a customer originating the question regardless of whether said determining step is positive or negative.

21. A computer system, comprising:

a processor; and

a memory coupled to said processor, the memory having stored therein sequences of instructions, which, when executed by said processor, causes said processor, after receiving a response to a question from a human agent, to perform the steps of;

using the human agent'"'"'s response, determining whether to add the response to the database;

if said determining step is positive then creating metadata for categorizing and describing said response, and updating the database with the response and said metadata; and

if said determining step is negative then rejecting the response.

22. A computer architecture, comprising:

receiving means for receiving a question and routing the question to a human agent;

determining means for;

receiving a response from said human agent to said question and determining whether to add the response to the database;

if said determining step is positive then creating metadata for categorizing and describing said response, and updating the database with the response and said metadata; and

if said determining step is negative then rejecting the response.

23. An article including a computer readable medium having stored thereon sequences of instructions, said plurality of sequences of instructions including sequences of instructions which, when executed by a processor cause said processor to perform the steps of:

receiving a question and routing the question to a human agent;

using a response from said human agent to said question, determining whether to add the response to the database;

if said determining step is positive then creating metadata for categorizing and describing said response, and updating the database with the response and said metadata; and

if said determining step is negative then rejecting the response.