System and Method for Automatically Generating Adaptive Interaction Logs from Customer Interaction Text

• US 20100104087A1
• Filed: 04/23/2009
• Published: 04/29/2010
• Est. Priority Date: 10/27/2008
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: International Business Machines Corporation
• Inventors: Gates, Stephen Carl, Neff, Mary S., Byrd, Roy J., Teiken, Wilfried, Park, Youngja
• Granted Patent: US 8,644,488 B2
• Time in Patent Office: Days
• US Class Current: 379/265.90
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: H04M 2201/40 using speech recognition
• CPC Class Codes: H04M 3/42221 Conversation recording systems at the subscriber's set H04M1/656
• CPC Class Codes: H04M 3/4936 Speech interaction details speech recognition per se G10L15/00
• CPC Class Codes: H04M 3/51 Centralised call answering arrangements requiring operator intervention , e.g. call or contact centers for telemarketing
• CPC Class Codes: H04M 3/5133 Operator terminal details
• CPC Class Codes: H04M 3/5191 interacting with the Internet

Abstract

A system and method for providing an adaptive Interaction Logging functionality to help agents reduce the time spent documenting contact center interactions. In a preferred embodiment the system uses a pipeline comprising audio capture of a telephone conversation, automatic speech transcription, text normalization, transcript generation and candidate call log generation based on Real-time and Global Models. The contact center agent edits the candidate call log to create the final call log. The models are updated based on analysis of user feedback in the form of the editing of the candidate call log done by the contact center agents or supervisors. The pipeline yields a candidate call log which the agents can edit in less time than it would take them to generate a call log manually.

Claims

1. A method for a computer apparatus for automatically generating a customer interaction log for an interaction between a customer and agent at a contact center comprising the steps of:

receiving input from at least one of said customer and said agent;

automatically analyzing said received input to generate a customer interaction log using at least one model;

displaying said customer interaction log for agent review at a graphical user interface of an agent computer;

receiving agent feedback at said graphical user interface; and

automatically updating at least one of said customer interaction log and said at least one model based on said agent feedback.

2. The method of claim 1 wherein said receiving input comprises at least one of:

receiving text of speech from a speech recognition system;

receiving text entered on a Web form;

receiving text of instant messaging; and

receiving text of emails.

3. The method of claim 2 further comprising a step of normalizing said received input by performing at least one of the steps of:

removing disfluencies from said received input;

normalizing vocabulary in said received input;

normalizing alphanumeric characters in said received input;

correcting misspellings or incorrectly recognized words in said received input;

detecting sentence boundaries in said received input;

capitalizing letters in sentences in said received input; and

detecting call segment boundaries in said received input.

4. The method of claim 1 wherein said automatically analyzing the received input to generate a customer interaction log further comprises at least one of:

scoring utterances in received input based on at least one model;

identifying selected utterances to be used in the customer interaction log; and

correcting words in selected utterances.

5. The method of claim 1 further comprising initial creation of said at least one model by at least one of:

manual rule creation; and

machine learning rule creation.

6. The method of claim 1 further comprising analyzing said feedback and updating of said at least one model based on feedback analysis.

7. The method of claim 6 wherein said determining updating comprises comparing an edited customer interaction log edited by the agent to the automatically generated customer interaction log.

8. The method of claim 1 wherein automatically updating said at least one model further comprises at least one of:

decreasing the salience of a rule which generated a deleted sentence;

increasing the salience of a rule which generated an inserted sentence;

finding a most similar sentence in the received input to a newly created sentence; and

increasing salience of a rule which generated the most similar sentence.

9. The method of claim 1 further comprising storing at least one of the customer interaction log and agent feedback.

10. The method of claim 9 wherein agent feedback is stored for a plurality of interactions between customers and agents and further comprising analyzing said stored feedback and determining updating of a Global Model based on said stored feedback analyzing.

11. The method of claim 1 wherein said customer interaction log is generated with reference to a RealTime Model automatically generated for said interaction.

12. The method of claim 11 wherein said RealTime Model is automatically generated from a Global Model.

14. The system of claim 12 further comprising a text normalizing component for transforming received input comprising at least one of:

a disfluency component removing disfluencies from the received text;

a vocabulary normalizing component for normalizing vocabulary in said received text;

a character normalizing component for normalizing alphanumeric characters in said received text;

a word correction component for correcting misspellings or incorrectly recognized words in said received text;

a sentence detection component for detecting sentence boundaries in said received text;

a sentence capitalization component for capitalizing letters in sentences in said received text; and

a call segment component for detecting call segment boundaries in said received text.

13. A system for automatically generating a customer interaction log for an interaction between a customer and agent at a contact center comprising:

at least one processing unit for executing components;

a receiving component for receiving input from at least one of the customer and the agent;

an analysis component for automatically analyzing the received input to generate a customer interaction log using at least one model;

a display generation component for generating a display of said customer interaction log for agent review at a graphical user interface of an agent computer;

a feedback collection component for receiving agent feedback; and

at least one learning component for analyzing agent feedback to determine updating of at least one of the customer interaction log and said at least one model based on agent feedback.

15. The system of claim 13 wherein said analysis component comprises at least one subcomponent for performing at least one of:

scoring utterances in received input based on at least one model; and

identifying selected utterances to be used in the customer interaction log.

16. The system of claim 13 further comprising at least one storage component for storing at least one of said at least one models, said customer interaction log and said agent feedback.

17. The system of claim 13 wherein said analysis component additionally comprises at least one of:

an utterance scoring component for scoring utterances in said call transcript based on at least one of a Global Model and a RealTime Model;

an utterance selection component for identifying selected utterances to be used in said interaction log; and

18. The system of claim 16 wherein agent feedback is stored for a plurality of interactions between customers and agents and wherein said at least one learning component analyzes said stored feedback and determines updating of a Global Model based on said stored feedback analyzing.

19. A program storage device readable by machine tangibly embodying a program of instructions executable by a processing device for performing a method for automatically generating a customer interaction log for an interaction between a customer and agent at a contact center for display to the agent comprising the steps of:

receiving input from at least one of the customer and the agent;

automatically analyzing the received input to generate a customer interaction log using at least one model;

displaying the customer interaction log for agent review at a graphical user interface of an agent computer;

receiving agent feedback at said graphical user interface; and

automatically updating at least one of the customer interaction log and said at least one model based on the agent feedback.