Interactive electronic messaging system

• US 6,477,551 B1
• Filed: 02/16/1999
• Issued: 11/05/2002
• Est. Priority Date: 02/16/1999
• Status: Expired due to Term

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Johnson, David E., Oles, Frank J.
• Primary Examiner(s): Feild, Joseph H.
• Application Number:
  US09/249,797
• Time in Patent Office: 1,358 Days
• Field of Search: 707/4, 707/9, 707/531, 707/5, 709/206, 345/752, 706/45, 704/1
• US Class Current: 715/202
• CPC Class Codes: G06F 40/284 Lexical analysis, e.g. tokenisation or collocates
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access

Abstract

A method and apparatus for interacting with incoming text information, e.g., a user query. The method of the present invention categorizes the incoming text information, and may also provide associated confidence levels with the categorization feature. The categorized information is used to query a database having associated category information with associated threshold values. The confidence levels may also be used when querying the database. The categorized information and corresponding confidence levels are then compared to the threshold value. If the categorized information and corresponding confidence levels equals or exceeds the threshold value then a response is provided by the system and method of the present invention. If the categorized information and corresponding confidence levels does not equal or exceed the threshold value then the system of the present invention will request further information of the user, and the method will repeat until (i) the categorized information and corresponding confidence levels equals or exceeds the threshold value or (ii) the system of the present invention times out.

Claims

1. A method of interactively generating a response to an incoming electronic message, comprising:

analyzing text information of the incoming electronic message to categorize the incoming text information into categorization information, the text information being a user input query;

retrieving category information from a data storage unit corresponding to the categorization information, the retrieved category information having an associated threshold value which is a predetermined value;

determining whether the categorization information equals or exceeds the threshold value associated with the retrieved category information; and

requesting further text information from the user when the categorization information does not equal or exceed the threshold value, wherein the request for further text information is machine generated.

2. The method of claim 1, further comprising:

providing the further text information in response to the requesting step, the further text information being analyzed with the text information to categorize the further text information and the text information into the categorization information; and

repeating the retrieving and determining steps.

3. The method of claim 2, further comprising responding to the input user query when the categorization information equals or exceeds the threshold value, the response being associated with the retrieved category information.

4. The method of claim 1, further comprising responding to the input user query when the categorization information equals or exceeds the threshold value, the response being associated with the retrieved category information.

5. The method of claim 1, wherein the text information is user input free-form text information, or structured or semi-structured text information.

6. The method of claim 1, wherein the analyzing step further comprises providing confidence level information associated with the categorization information, the confidence level information and the categorization information being used in the determining step, the confidence level information being derived by:

training on the text information by breaking the text information into tokens and counting the frequencies of the tokens;

providing a statistical analysis of the text information in order to provide a set of classification rules, and

applying the classification rules to the training data, where each document associated with the text information in the training data includes a known set of class labels.

7. The method of claim 6, wherein the training on the text information further includes morphological analysis, syntactic parsing, or statistical analysis.

8. The method of claim 1, wherein the requesting step requests further text information from the user when the user has not provided the further text information more than N times, where N is an integer, the further text information being analyzed with the text information to categorize the further text information and the text information into the categorization information.

9. The method of claim 8, further comprising providing the user with a template for providing structured or semi structured text in response to the requesting step.

10. The method of claim 1, further comprising providing the user with a template for providing structured or semi structured text in response to the requesting step.

11. The method of claim 1, further comprising timing out when the categorization information does not equal or exceed the threshold value and when the user has provided further text information more than N times, where N is an integer.

12. The method of claim 1, further comprising requesting further text information from the user when there is no retrieved category information stored within the data storage unit associated with the categorization information.

13. The method of claim 1, further comprising storing user information in a second storage unit, the stored user information being analyzed with the text information to categorize the stored user information and the incoming text information into the categorization information.

14. The method of claim 13, further comprising repeating the steps of claim 1 until the categorization information equals or exceeds the threshold value such that when the categorization information equals or exceeds the threshold value a machine generated response is provided to the user.

15. An article of manufacture comprising:

a computer usable medium having computer readable program code means embodied therein for analyzing text information of incoming electronic message to categorize the incoming text information into categorization information, the text information being a user input query;

a computer usable medium having computer readable program code means embodied therein for retrieving category information from a data storage unit corresponding to the categorization information, the retrieved category information having an associated threshold value which is a predetermined value;

a computer usable medium having computer readable program code means embodied therein for determining whether the categorization information equals or exceeds a threshold value associated with the retrieved category information; and

a computer usable medium having computer readable program code means embodied therein for requesting further text information from the user when the categorization information does not equal or exceed the threshold value, wherein the request for further information is machine generated.

16. The article of manufacture of claim 15, further comprising:

a computer usable medium having computer readable program code means embodied therein for providing the further text information in response to the generating step, the further text information being analyzed with the text information to categorize the incoming text information and the further text information into the categorization information; and

a computer usable medium having computer readable program code means embodied therein for responding to the input user query when the categorization information equals or exceeds the threshold value, the response being associated with the retrieved category information.

17. A means for interactively generating a response to an incoming electronic message, comprising:

means for analyzing text information of the incoming electronic message to categorize the incoming text information into categorization information, the text information being a user input query;

means for retrieving category information from a data storage unit corresponding to the categorization information, the retrieved category information having predetermined associated threshold values;

means for determining whether the categorization information equals or exceeds a threshold value; and

means for requesting further text information from the user when the categorization information does not equal or exceed the threshold value, wherein the request for further information is machine generated.

18. The means of claim 17, further comprising means for responding to the input user query when the categorization information equals or exceeds the threshold value, the response being associated with the retrieved category information.

19. The means of claim 17, wherein the means for analyzing further comprises means for providing confidence level information associated with the categorization information, the confidence level information being derived from:

means for training on the text information by breaking the text information into tokens and counting the frequencies of the tokens;

means for providing a statistical analysis of the text information to provide a set of classification rules, and

means for applying the classification rules to the training data.

20. The means of claim 18, wherein the means for requesting further text information from the user requests information from the user when the user has not provided further text information more than N times, where N is an integer.

21. The means of claim 17, further comprising means for timing out when the categorization information does not equal or exceed the threshold value and when the user has provided further text information more than N times, where N is an integer.

22. The means of claim 17, wherein means for requesting further text information from the user includes requesting the further text information when there is no retrieved category information stored within the data storage unit associated with the categorization information.

23. The means of claim 17, further comprising means for storing user information, the stored user information being analyzed with the text information of the incoming electronic message to categorize the user information and the incoming text information into the categorization information.

24. An apparatus that interactively generates a response to an incoming electronic message, comprising:

an analyzing module, the analyzing module categorizing the incoming text information into categorization information, the text information being a user input query;

a storing module associated with the analyzing module, the storing module being a database that stores category information corresponding to the categorization information obtained from the analyzing module;

a retrieving and determining module associated with the storing module, the retrieving and determining module retrieves the category information corresponding to the categorization information and determines whether the categorization information equals or exceeds a threshold value, the threshold value being a predetermined value associated with the stored category information; and

a requesting module associated with the retrieving and determining module, the requesting module requesting further text information from the user when the categorization information does not equal or exceed the threshold value, wherein the request for further information is machine generated.

25. The apparatus of claim 24, further comprising a responding module, the responding module responds to the input user query when the categorization information equals or exceeds the threshold value, the response being associated with the retrieved category information.

26. The apparatus of claim 24, wherein the analyzing module further provides confidence level information associated with the categorization information, the confidence level information being used in the retrieving and determining module with the categorization information, the confidence level information being derived from:

a training module training on the text information by breaking the text information into tokens and counting the frequencies of the tokens;

a statistical analysis module that provides a set of classification rules, and

means for applying the classification rules obtained from the statistical analysis module to the training data obtained from the training module.

27. The apparatus of claim 24, wherein the requesting module requests information from the user when the user has not provided further text information more than N times, where N is an integer.

28. The apparatus of claim 24, further comprising a timing out module for timing out when the categorization information of the analyzing module does not equal or exceed the threshold value and when the user has provided further text information more than N times, where N is an integer.

29. The apparatus of claim 24, wherein the requesting module requests the further text information when there is no retrieved category information stored within the storing module associated with the categorization information.

30. The apparatus of claim 24, further comprising a second storing module that stores user information and being associated with the analyzing module, the stored user information being analyzed in the analyzing module with the text information to categorize the stored user information and the incoming text information into the categorization information.