Method and system for processing messages within the framework of an integrated message system

• US 20060227945A1
• Filed: 10/14/2005
• Published: 10/12/2006
• Est. Priority Date: 10/14/2004
• Status: Active Grant

• Current Assignee: Deutsche Telekom AG
• Original Assignee: Deutsche Telekom AG
• Inventors: Liu, Jin, Runge, Fred, Trinkel, Marian, Mueller, Christel, Niedermueller, Frank, Schonebeck, Heiko-Armin
• Granted Patent: US 7,809,117 B2
• Time in Patent Office: Days
• US Class Current: 379/88.50
• CPC Class Codes: H04M 2201/40 using speech recognition
• CPC Class Codes: H04M 2201/60 Medium conversion
• CPC Class Codes: H04M 2203/2061 Language aspects
• CPC Class Codes: H04M 2203/303 Marking
• CPC Class Codes: H04M 2203/4509 Unified messaging with single point of access to voicemail and other mail or messaging systems
• CPC Class Codes: H04M 3/2218 Call detail recording
• CPC Class Codes: H04M 3/382 using authorisation codes or passwords
• CPC Class Codes: H04M 3/42059 Making use of the calling party identifier
• CPC Class Codes: H04M 3/5307 for recording messages comprising any combination of audio and non-audio components
• CPC Class Codes: H04M 3/5335 Message type or catagory, e.g. priority, indication
• CPC Class Codes: H04M 3/53358 Message preview

Abstract

A method and system for processing messages within the framework of an integrated message system. Recipients of messages in an integrated messaging system are provided with an authentic impression of the received message. In a first step, a message received within the framework of an integrated messaging system is automatically translated. Language detection and dictation system is provided. The message contents of the incoming message as well as its segments and parameters are simultaneously utilized to generate additional information regarding the sender and the information, which is suitable to give the recipient an impression of the received message in the most authentic form possible.

Claims

1. A method for processing messages within the framework of

an integrated messaging system, wherein, after ascertaining the identification, the text data and voice data to be gathered from the instantaneous message of a sender are processed by a translation system and, after processing, are analyzed together with

sound data, image data, messages with other types of media in connection with the instantaneous message, and data derived from additional information sources;

data that were stored within the framework of previous messages of the sender in a sender data area for acquired and/or ascertained sender data; and

data of additional features from the owner data area for acquired and/or ascertained owner data, such analysis being implemented according to semantic, prosodic, phonetic and other analytical methods, as well as image processing methods, for the presence of information/data that are suitable to convey to the recipient both the message itself and also the background information in connection with the particular message in the most authentic form possible, both the instantaneous message and the data derived from the message being evaluated according to predefined classification rules and assigned to different classes of a linking matrix configured as assignment matrix (n;

m), and the data provided within the framework of the linking matrix being stored in an index database and/or in a memory area for supplementary information of individual messages.

2. The method as recited in claim 1, wherein, to translate the instantaneous message, the language of the message and/or a segment of the message is determined according to the relation

L=arg max(W_CLI(Lx)+W_LD*C_LD(L_x)+W_T*C_T(L_x)+W_H*N_H(L_x)+ . . . +W_n*C_n(L_x))

the relation being implemented sequentially for each source language supported in the system, and the language (L_x) for which L exhibits the maximum value being considered the source language (L) as a result, and in the case of identically high maximum values or maximum values for at least two languages that are very similar, information for correcting the weighting factors being output automatically.

3. The method as recited in claim 1, wherein the data record containing the result of the language detection of the source language is analyzed both for the further adjustment with respect to the automatic selection of the source language for the translation system and for the automatic selection of the target language into which the voice and/or text massage is to be translated.

4. The method as recited in claim 1, wherein the control of the individual translation modules of the translation module with decision matrix is implemented via the decision matrix into which both the result of the detection of the source language and the ascertained target language are input automatically.

5. The method as recited in claim 1, wherein, in addition to the data resulting from the translation of the message, additional information is provided to the recipient of the message which is ascertained by utilizing principles of associate knowledge management.

6. The method as recited in claim 1, wherein, in addition to the data relating to the profile of the sender, the identification such as name, first name, sender identification e-mail address, country code, gender, different call numbers and role assignment for defined call numbers, the data record of the mailbox owner also stores data regarding source languages that are excluded from a translation.

7. The method as recited in claim 1, wherein, to ascertain the emotional state of the sender of a message, both the results of the prosodic analyses of voice messages and the results of the semantic analysis of attached text files, as well as available image and/or video information are being analyzed.

8. The method as recited in claim 1, wherein the gender and/or the age of the sender of the message is ascertained with the aid of data from at least one module for speaker classification and/or with the aid of data from a module for ascertaining the most probable gender while including data from earlier messages from the sender data area for acquired and/or ascertained sender data including supplementary information.

9. The method as recited in claim 1, wherein instantaneous data for the most probable emotional state of the sender of the message together with the data regarding age and gender of the sender of the message are used for further statistical analysis;

these data together with additional data ascertained by analyzing the sender information are combined into data records within the framework of parameters and/or parameter linkages, a division being implemented into static and/or dynamic parameters and/or parameter linkages.

10. The method as recited in claim 1, wherein the data regarding the most probable emotional state of the sender of the message is used to classify the messages in a priority list pertaining to the incoming messages.

11. The method as recited in claim 1, wherein the incoming messages are assigned to different message types, the data resulting from the different message types being processed according to predefined classification rules and assigned to different classes/categories via a linking matrix, and the data, having been assigned to the particular classes and/or categories, are assigned to an index database of the mailbox system operator and/or stored in corresponding classes in the memory area for supplementary information of individual messages of the mailbox of the individual mailbox owner.

12. The method as recited in claim 1, wherein the data stored in the index database of the mailbox system operator and/or the data stored in the memory area for supplementary information of individual information and sorted according to classes/categories are provided with coefficients within the classes/categories, which determine the reliability of the assignment to a specific class.

13. A system for processing messages within the framework of an integrated messaging system, comprising

at least one translation system with automatic foreign language detection;

at least one module for prosodic analysis of speaker emotions;

at least one module for semantic analysis of sender emotions;

at least one data memory with information regarding the languages that are able to be processed by the available emotion-analysis module and/or the translation module(s);

at least one data memory with sample data and/or dynamic/static parameters and/or parameter linkages for analyzing collected sender information;

at least one sender data area for acquired and/or ascertained sender data to which the data ascertained via the aforementioned module are supplied; and

at least classification block.

14. The system as recited in claim 13, wherein the translation system is made up of

a module for language identification/language detection with the aid of a text, having access to an ASR module with algorithms for language detection;

a language evaluation module connected to the module for language identification/language detection on the basis of a text, having access to a memory area for supplementary information of individual messages, to a module for language detection, and to a module for ascertaining the most probable gender;

a decision module connected to the module for language identification/language detection on the basis of a text, having outputs to the input module for source languages, to an interface module for selecting the source language and to a module for supplementing the messages by supplementary information, the decision module having access to the ASR module and being made up of

a translation module with decision matrix, which is linked to the input module for source languages having the source languages (S1-Sn), the output module for target languages having the target languages (S1-Sm), the interface module for selecting the source language, and a module for selecting the target language, the module for selecting the target language having a cross connection to the owner data area for acquired and/or ascertained owner data, and the output module for target languages having cross connections to the module for supplementing the messages by supplementary information, to a module for analyzing sender emotions and to a classification block.

15. The system as recited in claims 13 and 14, wherein the classification block to which all data including supplementary data of a message are supplied according to the data type, is made up of

a message classification module in which the messages are sorted according to their message type;

a linking matrix, which is connected to the message classification module and which is configured as assignment matrix (n;

m) for different classes/categories; and

an index database connected to the linking matrix, the message classification module having a cross connection to the memory for static and dynamic parameter linkages for analyzing sender and owner information, to the data memory with sample data and to the output module for target languages, and the linking matrix and the index database having a cross connection to the mailbox.