Optimizing a language/media translation map

• US 8,364,465 B2
• Filed: 03/19/2012
• Issued: 01/29/2013
• Est. Priority Date: 09/25/2009
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Miyamoto, Kohtaroh, Sobhi, Ali
• Primary Examiner(s): He, Jialong
• Application Number:
  US13/423,537
• Publication Number:
  US 20120179451A1
• Time in Patent Office: 316 Days
• Field of Search: 704/2, 704/3, 704/4, 704/8
• US Class Current: 704/2
• 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 15/26 Speech to text systems G10L15/08 takes precedence

Abstract

A mechanism is provided for optimizing a language/media translation map. A user input is received comprising an input language/media selection, one or more output languages/medias selections, and a threshold for at least one of accuracy or throughput of one or more requested language/media translations. For each of the one or more requested language media translations, a determination is made as to whether an accuracy or throughput of a selected one of an automated translation system or a human resource translator is above the threshold for the at least one of accuracy or throughput. Responsive to the accuracy or throughput being above the threshold, either the selected one of the automated translation system or the selected one of the human resource translator is added to a multiple language/media translation map. An optimized multiple language/media translation map is then generated for use by a translation orchestration module in the data processing system.

Claims

1. A method, in a data processing system, for optimizing a language/media translation map, the method comprising:

receiving, by a language/media translation map optimization module of the data processing system, user input comprising an input language/media selection, one or more output languages/medias selections, and a threshold for at least one of accuracy or throughput of one or more requested language/media translations;

for each of the one or more requested language media translations, determining, by the language/media translation map optimization module, whether an accuracy or throughput of a selected one of an automated translation system or a human resource translator is above the threshold for the at least one of accuracy or throughput;

responsive to the accuracy or throughput of the selected one of the automated translation system or the human resource translator being above the threshold for the at least one of accuracy or throughput, adding, by the language/media translation map optimization module, either the selected one of the automated translation system or the selected one of the human resource translator to a multiple language/media translation map;

generating, by the language/media translation map optimization module, an optimized multiple language/media translation map utilized by a translation orchestration module in the data processing system;

determining by the language/media translation map optimization module, whether a selected human resource translator has been assigned to more than one of the one or more requested language/media translations;

responsive to the selected human resource translator being assigned more than one translation, determining, by the language/media translation map optimization module, whether there is another human resource translator or another automatic translation system to perform one of the more than one assigned translations with an accuracy and/or throughput above the predefined threshold;

responsive to the existence of the another human resource translator or automatic translation system to perform one of the more than one assigned translations with an accuracy and/or throughput above the predefined threshold, selecting, by the language/media translation map optimization module, either the other human resource translator or the other automatic translation system to perform the translation; and

adding, by the language/media translation map optimization module, either the other automated translation system or the other human resource translator to perform the one or more intermediate translations to the multiple language/media translation map.

2. The method of claim 1, further comprising:

for each of the one or more requested language media translations, determining, by the language/media translation map optimization module, whether each requested language/media translation of the one or more requested language media translations may be performed directly;

responsive to a failure to perform the requested language/media translation directly, determining, by the language/media translation map optimization module, whether one or more intermediate translations may be performed in order to achieve the requested language/media translation;

responsive to identifying one intermediate translation to be performed, selecting, by the language/media translation map optimization module, the one intermediate translation for the requested language/media translation, thereby forming two required translations to produce the requested language/media translation;

for a first translation of the two required translations, replacing, by the language/media translation map optimization module, the one or more output language/media selections with a language/media of the one intermediate translation;

for a second of the two required translations, replacing, by the language/media translation map optimization module, the input language/media with the language/media of the one intermediate translation; and

adding, by the language/media translation map optimization module, either the selected one of the automated translation system or the selected one of the human resource translator to perform each of the two required translations to the multiple language/media translation map.

3. The method of claim 2, further comprising;

responsive to a failure to identity one or more intermediate translations that may be performed, sending, by the language/media translation map optimization module, an error to a user.

4. The method of claim 1, further comprising:

responsive to the accuracy or throughput of the selected one of the automated translation system or the human resource translator being below the threshold for the at least one of accuracy or throughput, determining, by the language/media translation map optimization module, whether one or more intermediate translations may be performed in order to achieve each of the one or more requested language/media translations;

responsive to identifying one intermediate translation to be performed, selecting, by the language/media translation map optimization module, the one intermediate translation for the requested language/media translations, thereby forming two required translations to produce the requested language/media translations;

for a first translation of the two required translations, replacing, by the language/media translation map optimization module, the one or more output language/media selections with a language/media of the one intermediate translation;

for a second of the two required translations, replacing, by the language/media translation map optimization module, the input language/media with the language/media of the one intermediate translation; and

adding, by the language/media translation map optimization module, either the selected one of the automated translation system or the selected one of the human resource translator to perform each of the two required translations to the multiple language/media translation map.

5. The method of claim 4, further comprising:

responsive to a failure to identify one or more intermediate translations that may be performed, noting, by the language/media translation map optimization module, an exception to a user; and

adding, by the language/media translation map optimization module, either the automated translation system or the human resource translator to the multiple language/media translation map.

6. The method of claim 1, further comprising:

responsive to a failure to identify another human resource translator or automatic translation system to perform one of the more than one assigned translations with an accuracy and/or throughput above the predefined threshold, adding, by the language/media translation map optimization module, either the selected one of the automated translation system or the selected one of the human resource translator to perform the one or more intermediate translations to the multiple language/media translation map.

7. The method of claim 1, further comprising:

incrementing for each of the one or more requested language/media translations, by the language/media translation map optimization module, a cost counter by predefined cost associated with each of the selected one of the automatic translation system, the selected one of the human resource translator, the other automated translation system, or the other human resource translator;

responsive to all costs for each of the one or more requested language/media translations being added to the cost counter, determining, by the language/media translation map optimization module, whether a value of the cost counter exceeds a predefined budget for the one or more requested language/media translations; and

responsive to the value of the cost counter exceeding the predefined budget, sending, by the language/media translation map optimization module, a notification of a surpass to a requestor of the one or more requested language/media translations.