Systems and methods for translation management

Patent metadata

Field	Value
Publication number	US-10803256-B2
Application number	US-201715858156-A
Country	US
Kind code	B2
Filing date	Dec 29, 2017
Priority date	Dec 29, 2017
Publication date	Oct 13, 2020
Grant date	Oct 13, 2020

Abstract

Systems and methods for a translation management system include performing a source text collection and translation process. The source text collection and translation process includes collecting, from one or more applications, one or more source texts for translation. Source segments for translation are determined using the one or more source texts. Source text properties associated with the one or more source texts are provided to a machine learning engine. Translation performance requirement predictions associated with the plurality of source segments respectively are generated by the machine learning engine based on the source text properties. A plurality of translation requests associated with the plurality of source segments is provided by the machine learning engine based on the translation performance requirement predictions. One or more translated texts generated in response to executing the plurality of translation requests are received. A translation result storage is updated using the one or more translated texts.

First claim

What is claimed is:

1. A system for performing source text collection and translation, comprising:
a non-transitory memory storing instructions; and
one or more hardware processors coupled to the non-transitory memory and configured to read the instructions from the non-transitory memory to cause the system to perform operations comprising:
collecting periodically, from one or more applications, one or more source texts for translation, wherein the one or more source texts are under development and are updates from one or more previously translated source texts;
providing, to a machine learning engine, source text properties associated with the one or more source texts;
generating, by the machine learning engine, a plurality of translation performance requirement predictions associated with a plurality of source segments of the one or more source texts respectively based on the source text properties;
based on the plurality of translation performance requirement predictions, generating, by the machine learning engine, a plurality of translation requests associated with the plurality of source segments for translation of the one or more source texts;
determining that the plurality of translation requests associated with the plurality of source segments exceeds a total daily word count limit determined by the machine learning engine based on a translation capacity learned by the machine learning engine to flatten a translation demand associated with a development time period of the one or more source texts under development;
in response to the determining that the plurality of translation requests associated with the plurality of source segments exceeds the total daily word count limit, providing on a future day, by the machine learning engine, the plurality of translation requests associated with the plurality of source segments;
receiving one or more translated texts generated in response to executing the plurality of translation requests; and
updating a translation result storage using the one or more translated texts.

2. The system of claim 1 , wherein the one or more applications are provided by a development system provider device; and
wherein the operations further comprise:
repeating the collecting periodically, from the one or more applications, the one or more source texts for translation, at a first frequency;
receiving a first production request associated with a first application of the one or more applications; and
in response to receiving the first production request, providing first translated texts associated with the first application for localizing the first application by a production system provider device.

3. The system of claim 1 , wherein a first source text property of the source text properties associated with a first source text includes an importance level of a page containing the first source text.

4. The system of claim 1 , wherein a first source text property of the source text properties associated with a first source text includes user usage information of a page containing the first source text.

5. The system of claim 1 , wherein the generating the plurality of translation performance requirement predictions is further based on source segment properties associated with the plurality of source segments.

6. The system of claim 5 , wherein a first source segment property associated with a first source segment includes a property selected from the group consisting of a word count, a terminology, a grammar complexity level, and a grammar difficulty level.

7. A method, comprising:
collecting periodically, from one or more applications, one or more source texts for translation, wherein the one or more source texts are under development and are updates from one or more previously translated source texts;
providing, to a machine learning engine, source text properties associated with the one or more source texts;
generating, by the machine learning engine, a plurality of translation performance requirement predictions associated with a plurality of source segments of the one or more source texts respectively based on the source text properties;
based on the plurality of translation performance requirement predictions, generating, by the machine learning engine, a plurality of translation requests associated with the plurality of source segments for translation of the one or more source texts;
determining that the plurality of translation requests associated with the plurality of source segments exceeds a total daily word count limit determined by the machine learning engine based on a translation capacity learned by the machine learning engine to flatten a translation demand associated with a development time period of the one or more source texts under development;
in response to the determining that the plurality of translation requests associated with the plurality of source segments exceeds the total daily word count limit, providing on a future day, by the machine learning engine, the plurality of translation requests associated with the plurality of source segments;
receiving one or more translated texts generated in response to executing the plurality of translation requests; and
updating a translation result storage using the one or more translated texts.

8. The method of claim 7 , further comprising:
generating, by the machine learning engine, a plurality of translator performance predictions associated with a plurality of translators respectively, wherein a first parameter of a first translation request includes a first translator for performing the first translation request.

9. The method of claim 8 , wherein the first translator is a machine translation provider.

10. The method of claim 8 , wherein the first translator is a human translator.

11. The method of claim 8 , wherein a translation process parameter of the first translation request includes a first number of review cycles based on the plurality of translation performance requirement predictions and the plurality of translator performance predictions.

12. The method of claim 11 , further comprising:
receiving, by the machine learning engine, feedback from the one or more review cycles; and
performing a training process, to the machine learning engine, based on the feedback.

13. A non-transitory machine-readable medium having stored thereon machine-readable instructions executable to cause a machine to perform operations comprising:
collecting periodically, from one or more applications, one or more source texts for translation, wherein the one or more source texts are under development and are updates from one or more previously translated source texts;
providing, to a machine learning engine, source text properties associated with the one or more source texts;
generating, by the machine learning engine, a plurality of translation performance requirement predictions associated with a plurality of source segments of the one or more source texts respectively based on the source text properties;
based on the plurality of translation performance requirement predictions, generating, by the machine learning engine, a plurality of translation requests associated with the plurality of source segments for translation of the one or more source texts;
determining that the plurality of translation requests associated with the plurality of source segments exceeds a total daily word count limit determined by the machine learning engine based on a capacity learned by the machine learning engine to flatten a translation demand associated with a development time period of the one or more source texts under development;

in response to the determini

Assignees

• Paypal Inc

Inventors

• Lagunas Diego

Classifications

• G06F40/51 Primary

  Translation evaluation · CPC title

• G06N20/00

  Machine learning · CPC title

• G06F40/58

  Use of machine translation, e.g. for multi-lingual retrieval, for server-side translation for client devices or for real-time translation · CPC title