Communication system using mixed translating while in multilingual communication

• US 8,706,471 B2
• Filed: 05/18/2007
• Issued: 04/22/2014
• Est. Priority Date: 05/18/2006
• Status: Active Grant

• Current Assignee: University of Southern California
• Original Assignee: University of Southern California
• Inventors: Narayanan, Shrikanth, Georgiou, Panayiotis
• Primary Examiner(s): Kazeminezhad, Farzad
• Application Number:
  US11/750,926
• Publication Number:
  US 20080071518A1
• Time in Patent Office: 2,531 Days
• Field of Search: 704/2, 364/419
• US Class Current: 704/2
• CPC Class Codes: G06F 40/295 Named entity recognition
• CPC Class Codes: G06F 40/55 Rule-based translation
• CPC Class Codes: G06F 40/58 Use of machine translation, e.g. for multi-lingual retrieval, for server-side translation for client devices or for real-time translation
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue

Abstract

A translation between a source language and a target language is disclosed. The source language items are divided, with primary and secondary source language items or named entities being identified, where the primary and secondary source languages being different from each other and from the target language. The entities in the second source language are translated in a different way. For example, they may be copied into the target language, or translated in a special way that is based on their meaning, e.g, into a term that has a more descriptive meaning in the target language.

Claims

1. A method comprising:

receiving, at a computing system, speech information to be translated from a primary source language to a target language, wherein the speech information includes a phrase;

segmenting, by the computing system, the phrase into a first part and a second part, wherein the second part includes a named entity;

translating, by the computing system, the first part of the phrase into a first translated portion in the target language; and

translating, by the computing system, the second part of the phrase into a second translated portion in the target language;

wherein translating the second part comprises

translating the second part in a first case by copying the second part into the target language when the second part is in a secondary source language different from the primary source language,

translating the second part in a second case into the target language when the second part has a corresponding translation in the target language, and

translating the second part in a third case into a generic word in the target language when the named entity of the second part has a first meaning that represents the named entity in the target language, the named entity has a second meaning when translated as a non-named entity into the target language, the first meaning is different than the second meaning, and the second meaning does not represent the named entity.

2. The method of claim 1, wherein the named entity in the first case is a proper name.

3. The method of claim 2, wherein the copying comprises repeating the secondary source language after voice morphing into the target language.

4. The method of claim 1, wherein the primary source language is English, the secondary source language is Spanish, and the target language is Arabic.

5. The method of claim 1, wherein the copying comprises using automated speech reading.

6. A computer-implemented system, comprising:

an interface that receives speech information for translation into a target language, wherein the speech information comprises a first input utterance;

and a processor programmed to;

identify a non-named entity in a first portion of the first input utterance and a named entity in a second portion of the first input utterance

wherein the first portion of the first input utterance is in a primary source language different than the target language, and the second portion of the first input utterance is in a secondary source language different from the primary source language and the target language; and

translate the first portion of the first input utterance into a first portion in the target language;

convey the second portion of the first input utterance, into a second portion in the target language;

and provide an output utterance comprising the first and second portions.

7. The system of claim 6, wherein:

the speech information comprises a second input utterance;

the processor is programmed to identify a non-named entity in a first portion of the second input utterance and a named entity in a second portion of the second input utterance, translate the first portion of the second input utterance into a first translated portion in the target language, translate the second portion of the second input utterance into one or more generic terms in a second translated portion in the target language, wherein the second translated portion has a first meaning that represents the named entity in the target language, the named entity has a second meaning when translated as a non-named entity into the target language, the first meaning is different than the second meaning, and the second meaning does not represent the named entity; and

the processor is programmed to provide an output utterance comprising the first and second translated portions.

8. The system of claim 7, wherein the one or more generic terms include one or more non-named entities.

9. The system of claim 6, wherein the named entity in the second portion of the speech information includes a proper name.

10. The system of claim 6, wherein the processor is programmed to convey the second portion of the first input utterance into the second portion in the target language by repetition after voice morphing.