Method and system for adapting interactions

• US 9,183,833 B2
• Filed: 11/21/2007
• Issued: 11/10/2015
• Est. Priority Date: 11/22/2006
• Status: Active Grant

• Current Assignee: Deutsche Telekom AG
• Original Assignee: Deutsche Telekom AG
• Inventors: Runge, Fred, Johannsen, Wiebke, Oberle, Frank, Van Ballegooy, Markus, Burkhardt, Felix, Stegmann, Joachim, Eckert, Martin, Englert, Roman
• Primary Examiner(s): TILLERY, RASHAWN N
• Application Number:
  US11/943,878
• Publication Number:
  US 20080155472A1
• Time in Patent Office: 2,911 Days
• Field of Search: 715/705, 715/707, 715/708, 715/727, 715/728, 715/764, 715/765, 715/781, 715/788, 715/789, 715/810, 715/811, 715/978
• US Class Current: 1/1
• CPC Class Codes: G06F 40/35 Discourse or dialogue representation
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue

Abstract

A method and system for adapting automated interactions that allows the interactive behavior of an automated system, or the nature of the interaction elements implemented thereon, to be adapted to properties and/or behaviors of users of such systems in order to enhance operating convenience. Interaction adaptation is performed with reference to user groups to which the users are allocated.

Claims

1. A method for adapting interactions conducted between interaction systems and users of the interaction systems to the users, comprising the steps of:

defining a nature of a respective automated interaction by interaction states in the form of messages and prompts, by the configuration of the messages and prompts, by menu options and by state transitions; and

adapting the nature of the interaction, during its execution, to the user in a manner that is determined by a user class that is allocated, to the user conducting the interaction according to rules stored and implemented in the interaction system on the basis of properties communicated by the user during the interaction or automatically ascertained by the interaction system or the user'"'"'s behavior, or on the basis of the user'"'"'s identification by the system and information previously acquired with regard to the user'"'"'s identity,

wherein a manner in which the nature of the interaction is adapted to the user of a user class is modifiable,

wherein user-class-specific utilization counters for selected state transitions of a system-side part of the interaction are operated in the interaction system and are incremented or decremented upon passage through respective state transitions during processing of the interaction and when a respective counter, a group of counters or a variable derived from the counters, exceeds or falls short of an established limit value, an adaptation process is started with which the nature of the system-side part of the interaction is modified in user-class-related fashion, so that proceeding from an initial nature existing upon its implementation in the system, the system-side part of the interaction is repeatedly adapted in user-class-selective fashion to the needs of the members of the respective user class, and

wherein the user of the interaction system is allocated to one user class in terms of execution of the interaction that the user is presently conducting, wherein for the purpose of repeated modification of the manner in which the interaction is adapted during execution to persons of a user class, the user is allocated, as necessary, simultaneously to multiple user classes such that, during execution of the interaction, the nature of the interaction is repeatedly modified based on contemporaneous online interactions of a plurality of users of the multiple user classes.

2. The method according to claim 1 wherein a logical sequence of the state transitions determines a menu width and depth.

3. The method according to claim 1, wherein the user is allocated to only one user class in terms of execution of the interaction currently being conducted by the user, and is allocated, as necessary, to varying user classes in the course of the interaction.

4. The method according to claim 1, wherein user classes are also formed by combining other user classes that continue to be independently managed by the interaction system.

5. The method according to claim 1, wherein new user classes are formed automatically by the interaction system according to rules stored therein.

6. The method according to claim 1, wherein the interaction is implemented, for various user classes that are established or managed in the interaction system at the time of its implementation, using different user-class-specific initial natures.

7. The method according to claim 1, wherein the interaction is adapted in terms of its nature being embodied, at least partially, as a speech interaction, at least with regard to the user-side part of the interaction, allocation of the user to user classes of the interaction system accomplished on the basis of the language spoken by the user and detected by an arrangement for language identification.

8. The method according to claim 7, wherein allocation of the user to user classes of the interaction system is accomplished according to rules stored in the interaction system, by evaluating data automatically acquired by an arrangement for speech recognition.

9. The method according to claim 7, wherein allocation of the user to user classes of the interaction system is accomplished on the basis of an accent, dialect, or sociolect spoken by the user and identified by an arrangement for speaker classification.

10. The method according to claim 7, wherein allocation of the user to user classes of the interaction system is accomplished on the basis of the user'"'"'s age, determined with reference to the speech by the arrangement for speaker classification.

11. The method according to claim 1, wherein the interaction is adapted in terms of its nature being embodied at least partially as a speech interaction at least with regard to the user-side part of the interaction, allocation of the user to user classes of the interaction system accomplished on the basis of the user'"'"'s gender, determined with reference to the speech by the arrangement for speaker classification.

12. The method according to claim 1, wherein allocation of the user to user classes of the interaction system is accomplished automatically according to rules stored in the interaction system, by evaluating data automatically acquired by an arrangement for emotion detection.

13. The method according to claim 1, wherein at least one of properties and behaviors of the user are automatically ascertained by the interaction system and are linked to confidence values, and at least one of corresponding properties and behaviors are not taken into consideration in the event of an insufficient confidence for allocation of the user to user classes of the interaction system.

14. The method according to claim 1, wherein data used for allocation of the user to user classes of the interaction system includes properties or behaviors ascertained, or information previously acquired, with regard to the user'"'"'s identity, wherein the data used for allocation of the user to user classes is weighted so that individual items of the data take priority over others upon classification of the user in terms of the user'"'"'s allocation to user classes of the interaction system.

15. The method according to claim 14, wherein confidence values linked to the data used to classify the user affect the weight of the corresponding data.

16. The method according to claim 1, wherein the counters for the selected state transitions are weighted so that the respective counter values of the counters affect a user-class-related modification of the nature of the interaction with different weights.

17. The method according to claim 1, wherein allocation of the user to user classes of the interaction system is accomplished by combining at least one of properties and behaviors known or ascertained about the user with at least one of time-related, local, and technical parameters.

18. An automated interaction system, comprising:

at least one control and processing unit configured to adapt interactions conducted between an interaction system and users of the interaction system to the users, wherein a nature of a respective automated interaction is defined by interaction states in the form of messages and prompts, by the configuration of said messages and prompts, by menu options and by state transitions, wherein the nature of the interaction is adapted, during its execution, to the user in a manner that is determined by a user class that is allocated to the user conducting the interaction according to rules stored and implemented in the interaction system on the basis of properties communicated by the user during the interaction or automatically ascertained by the interaction system or the user'"'"'s behavior, or on the basis of the user'"'"'s identification by the system and information previously acquired with regard to the user'"'"'s identity, wherein the manner in which the nature of the interaction is adapted to the user of a user class is modifiable, wherein user-class-specific utilization counters for selected state transitions of a system-side part of the interaction are operated in the interaction system and are incremented or decremented upon passage through respective state transitions during processing of the interaction and when a respective counter, a group of counters or a variable derived from the counters, exceeds or falls short of an established limit value, an adaptation process is started with which the nature of the system-side part of the interaction is modified in user-class-related fashion, so that proceeding from an initial nature existing upon its implementation in the system, the system-side part of the interaction is repeatedly adapted in user-class-selective fashion to the needs of the members of the respective user class,

the control and processing unit including an input recognition device and an input evaluation device, at least one memory, an input device configured for interactive inputs of a user and inputs of an administrator, and an output device configured for outputting system communications and prompts, data held in the at least one memory of the interaction system including a user class manager,

wherein the control and processing unit is configured to manage the user-class-specific utilization counters for selected state transitions of the system-side part of the interactions imaged in the interaction system and the control and processing unit manages a control mechanism according to which, when values allocated to the counters, to a group of counters, or to variables derived therefrom exceed or fall below limit values, the system-side part of the interaction, proceeding from an initial nature existing upon its implementation in the system, is repeatedly adapted in user-class-selected fashion to needs of the members of a respective user class using a software program processed by the control and processing unit, without additional administrative interventions, and

wherein the control and processing unit is further configured to allocate the user of the interaction system to one user class in terms of execution of the interaction that the user is presently conducting and, for the purpose of repeated modification of the manner in which the interaction is adapted during execution to persons of a user class, the user is allocated, as necessary, simultaneously to multiple user classes such that, during execution of the interaction, the nature of the interaction is repeatedly modified based on contemporaneous online interactions of a plurality of users of the multiple user classes.

19. The interaction system according to claim 18 wherein the transitions between the interaction states and the transitions logical sequence determine a menu width and depth.

20. The interaction system according to claim 18, further comprising a language unit configured to provide language identification.

21. The interaction system according to claim 18, further comprising a speech recognition and evaluation unit.

22. The interaction system according to claim 18, further comprising:

a speaker unit configured to provide at least one of speaker identification and speaker classification; and

an evaluation unit associated with the speaker unit.

23. The interaction system according to claim 18, further comprising a handwriting recognition unit.

24. The interaction system according to claim 18, further comprising an emotion recognition and evaluation unit.