Knowledge-base system and method

• US 20020026435A1
• Filed: 08/16/2001
• Published: 02/28/2002
• Est. Priority Date: 08/26/2000
• Status: Abandoned Application

• Current Assignee: Interactive Intelligence Incorporated (Genesys Cloud Services Incorporated)
• Original Assignee: Interactive Intelligence Incorporated (Genesys Cloud Services Incorporated)
• Inventors: Hockema, Stephen Andrew, Brown, Donald Eugene, Wyss, Felix Immanuel
• Application Number:
  US09/931,209
• Publication Number:
  US 20020026435A1
• Time in Patent Office: Days
• US Class Current: 707/1
• CPC Class Codes: G06F 16/38 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 9/453 Help systems
• CPC Class Codes: G06Q 10/107 Computer-aided management of electronic mailing [e-mailing]
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G09B 7/00 Electrically-operated teaching apparatus or devices working with questions and answers mechanically operated G09B3/00; computing arrangements G06F

Abstract

A knowledge-base system includes a processor, a database, and a matcher for matching questions of clients to answers in the database. The knowledge-base system can be administered by emails sent to the system. Question-answer entries can be added, deleted, and modified through the use of emails and/or other forms. The system can be accessed over a network or directly accessed with an administration computer. Questions and answers are separately indexed in the system to increase the accuracy of generated results. The generated response sent to the client can be modified based on the question asked, the results, and predefined templates. The system can also detect when the system receives a reply from a customer in response to sent results. Once a reply is detected, the email is forwarded to an alternate email address.

Claims

1. A method, comprising:

operating a knowledge-base system configured to store a database containing answers to questions, the knowledge-base system being operatively coupled to a client computer;

receiving a message from the client computer;

determining the message from the client computer was a reply to a previously generated message from the knowledge-base system; and

forwarding the message from the client computer to a representative in response to said determining.

2. The method of claim 1, wherein the message from the client computer is an email.

3. The method of claim 1, further comprising:

enclosing a message identification number on all communications sent by the knowledge-base system; and

wherein said determining includes searching the message from the client computer for the message identification number.

4. The method of claim 3, wherein said forwarding includes attaching message history information.

5. The method of claim 1, further comprising:

maintaining a communication log of communications sent and received with the knowledge-base system; and

wherein said determining includes ascertaining with the communication log whether a reply detection limit has been exceeded for the client computer.

6. The method of claim 5, wherein the reply detection limit includes a communication interval limit of time intervals between successive communications with the client computer and a number of communications limits based on a number of communications with the client computer.

7. The method of claim 5, wherein the communications maintained in the communication log include email communications, web chatting communications, telephone conversations, and personal conversations.

8. The method of claim 1, wherein the knowledge-base system and the client computer are operatively coupled over a network.

9. The method of claim 1, further comprising:

enclosing a message identification number on all communications sent by the knowledge-base system;

maintaining a communication log of communications sent and received with the knowledge-base system; and

wherein said determining includes searching the message from the client computer for the message identification number and ascertaining with the communication log whether a reply detection limit has been exceeded for the client computer.

10. The method of claim 1, wherein the representative is a human being.

11. The method of claim 1, wherein the representative is an automated system.

13. The method of claim 12, wherein said creating includes modifying the response message to indicate that no answers were found when the query result does not contain any answers.

14. The method of claim 12, wherein said creating includes modifying the at least one of the response templates based on the question from the client computer.

15. The method of claim 12, wherein said creating includes selecting the at least one of the templates based upon the query result.

16. The method of claim 12, wherein said generating includes:

calculating a score for each question and answer in the database; and

selecting at least one answer based upon a dynamic threshold resulting from a statistical distribution of all scores.

18. The method of claim 17, wherein said evaluating includes weighting the answers more than the questions.

19. The method of claim 17, wherein said evaluating includes weighting the answers less than the questions.

20. The method of claim 17, further comprising generating one or more word indices, said selecting including referencing the word indices.

21. The method of claim 17, wherein said selecting includes scoring a length of one of the questions in proportion to a length of the query input.

22. The method of claim 17, wherein said selecting includes scoring the question-answer sets to create a distribution of scores and determining the query result based upon variability of the scores.

23. The method of claim 17, further comprising designating one or more words to ignore.

24. The method of claim 17, further comprising alerting a representative when a particular one of the question-answer sets is included in the output.

25. The method of claim 17, further comprising:

designating corresponding weights for the word in the questions and the answers; and

wherein said selecting includes scoring each of the questions and the answers using the corresponding weights.

26. The method of claim 17, further comprising defining aliases for at least one word.

27. The method of claim 17, further comprising designating one of the question-answer sets to always appear in the output for the query input.

28. The method of claim 17, further comprising designating one of the question-answer sets to never appear in the output for the query input.

29. The method of claim 17, further comprising proportionally weighting one of the question-answer sets to reduce likelihood of appearing in the output for the query input.

30. The method of claim 17, further comprising proportionally weighting one of the question-answer sets to increase likelihood of appearing in the output for the query input.

31. The method of claim 17, further adding a question-answer set to the database by email.

32. The method of claim 17, further comprising:

receiving a message from a client computer coupled to the system;

determining the message from the client computer was a reply to a previously generated message from the system; and

forwarding the message from the client computer to a representative in response to said determining.

33. The method of claim 32, further comprising:

creating a response message based on the message from the client computer, at least one of a number of response templates and the selected one or more sets, the system being configured to store the response templates each providing a different response format; and

wherein said providing the output includes sending the response message to the client computer.

35. The method of claim 34, wherein said determining includes:

calculating variability of scores from said scoring; and

basing the threshold limit on the variability of the scores.

36. The method of claim 34, further comprising:

defining an absolute threshold limit that is universally applicable to all received questions; and

selecting the question-answer sets with scores above the absolute threshold limit.

37. The method of claim 34, further comprising:

creating a reflexive index that includes the question from the client computer and at least the selected question-answer sets from said selecting;

scoring each question from the selected question-answer sets against the reflexive index;

scoring the question from the client against the reflexive index to generate a question score; and

choosing the selected question-answer sets with scores that favorably correlate with the question score.

38. The method of claim 37, wherein the reflexive index further includes all of the question-answer sets.

40. The method of claim 39, wherein the reflexive index further includes all of the question-answer sets.

12. A method, comprising:

operating a knowledge-base system configured to store a database containing answers to questions and a number of response templates each providing a different response format, the system being operatively coupled to a client computer;

receiving from the client computer an input corresponding to a question;

generating a query result from the database in response to the question from the client computer;

creating a response message based on the question from the client computer, at least one of the response templates, and the query result; and

sending the response message to the client computer.

17. A method, comprising:

operating a system configured with a database including a plurality of question-answer sets;

receiving a query input to the system, the query input including a word;

selecting one or more of the question-answer sets with the system in response to the query input by evaluating presence of the word in one or more answers of the question-answer sets differently than presence of the word in one or more questions of the question-answer sets; and

providing an output from the system based on said selecting.

34. A method, comprising:

operating a knowledge-base system configured to store a database formatted with a number of question-answer sets, the knowledge-base system being operatively coupled to a client computer;

receiving an input corresponding to a question from the client computer;

scoring the question-answer sets with respect to the question;

determining a threshold limit based upon said scoring; and

selecting the question-answer sets with scores above the threshold limit.

39. A method, comprising:

operating a knowledge-base system configured to store a database formatted with a number of question-answer sets, the knowledge-base system being operatively coupled to a client computer;

receiving an input corresponding to a question from the client computer;

selecting one or more candidate sets from the question-answer sets based on the question from the client computer;

creating a reflexive index that includes the question from the client computer and at least the candidate sets;

scoring each question from the candidate sets against the reflexive index;

scoring the question from the client against the reflexive index to generate a question score; and

choosing the candidate sets with scores that correlate with the question score.

41. A system comprising:

means for determining a message from a client computer was a reply to a previously generated message from a FAQ database and forwarding the message to a representative in response;

means for evaluating question components and answer components of the FAQ database independently relative to an input query; and

means for providing a response to the FAQ database query in accordance with one or more response templates, the response templates each relating to a different response format.

42. The system of claim 41, further comprising means for selectively hiding or ignoring words included in one or more question-answer entries of the FAQ database.

43. The system of claim 41, further comprising means for providing an alias designation.

44. A system for processing a question from a client computer, comprising:

a database formatted with a number of question-answer sets, said sets each including a question field and a corresponding answer field, each of said question and answer fields being configured to contain one or more words, said database further including a question index that associates each of said words in said question fields with each of said question-answer sets and an answer index that associates said words in said answer fields with each of said question-answer sets;

a matcher operatively coupled to said database to produce a query result using both said question index and said answer index in response to the question from the client computer; and

a processor operatively coupled to said matcher to send said query result to the client computer.