RWS Offline Language Translation

FTK Core and FTK Plus now allow users to integrate RWS Language Translation to enhance translation capabilities with machine learning. While the integration comes at a cost, it ensures users can see improvements in accuracy as well as processing times.


Support for FTK Central will be available soon.

Supported Languages

 

  • Albanian
  • Arabic
  • Armenian
  • Bengali
  • Bulgarian
  • Burmese
  • Catalan
  • Chinese (Simplified)
  • Chinese (Traditional)
  • Croatian
  • Czech
  • Danish
  • Dari
  • Dutch
  • English
  • Estonian
  • Finnish
  • French
  • Georgian
  • German
  • Greek
  • Hausa
  • Hebrew
  • Hindi
  • Hungarian
  • Indonesian
  • Italian
  • Japanese
  • Javanese

  • Khmer
  • Korean
  • Kurdish
  • Latvian
  • Lithuanian
  • Malay
  • Maltese
  • Norwegian
  • Pashto
  • Persian
  • Polish
  • Portuguese
  • Portuguese (Brazilian)
  • Romanian
  • Russian
  • Serbian
  • Slovakian
  • Somali
  • Spanish
  • Swahili
  • Swedish
  • Thai
  • Turkish
  • Ukrainian
  • Urdu
  • Uzbek
  • Vietnamese


Minimum System Requirements

RWS Language Weaver – Master Host

  • 1 CPU Core
  • 1GB RAM
  • 4GB Storage

Language Pair – Translation Engine

Actual requirements are proportional to the number of language pairs to be installed.

  • 1-4 CPU Cores [Increase core count for best performance]
  • 4GB RAM
  • 30GB Storage
  • 2GB GPU RAM [If utilising Neural Language Pairs ONLY]

Supported Operating Systems

Windows x64

Linux x64

Windows 10

CentOS 7

Windows 8.1

CentOS 8

Windows 8

Ubuntu 14.04

Windows 7 SP1

Ubuntu 16.04

Windows Server 2012

Ubuntu 18.04

Windows Server 2016

Ubuntu 20.04

Windows Server 2019

Red Hat Enterprise Linux (RHEL) 7

Red Hat Enterprise Linux (RHEL) 8

 

Firewall Rules

TCP Port

Descriptions

Required

8000

The Host/Server listens on the port for incoming client/browser HTTP(S) connections to the Language Weaver Edge WEB UI.

Yes

8001

The Host/Server listens on the port for incoming client HTTP(S) connections to the Language Weaver Edge REST API.

Yes

5671

The Host/Server listens on the port for incoming worker connections to the Language Weaver Edge Message Broker component.

Yes

5673

The Host/Server listens on the port for incoming worker connections to the Language Weaver Edge Message Broker component.

Yes

4637

The Host/Server listens on the port for incoming client connections to the Language Weaver Edge Host Agent API component.

Yes

Installation

While installing Language Weaver Edge, a domain-level service account with local administration rights is required in a distributed environment. 

The installer has two installation modes:

  • Master Host: Serves REST API, Web GUI, and optionally Job Engine(s) and/or Translation Engine(s)
  • Worker Host: Serves Job Engine(s) and/or Translation Engine(s)

Installing Language Weaver Edge – Master Host or Worker Host

  1. Locate the Language Weaver Edge installer and Run as Administrator.
  2. Click Next.
  3. Review and Accept to the License Agreement. Click Next.
  4. Select an Installation Directory and click Next.
  5. Select As a master host and click Next.
    • Select As a worker host if installing as a Translation Engine.
  6. Select No when prompt to install the embedded API components and click Next.
  7. Enter administrator details:
    • Username (Email Address)
    • Display Name
    • Password
  8. Click Next.
  9. Select Do not use HTTPS/TLS.
    • If required you can select either, Use user-provided certificates or Use self-signed certificates automatically generated by Language Weaver Edge.
  10. Click Next.
  11. Select the default server port. The default web server port is 8000.
  12. Click Next.
  13. Click Next again to start the installation process.
  14. When prompt, Allow Access for ets-initial-setup.exe.
  15. Click Finishwhen the installation is complete.
    • The Language Weaver Edge UI will appear in your default browser.
    • Enter the login credentials created during installation to test.

 

Installing Language Pairs

  1. Locate the Language Pair setup installer and Run as Administrator.
  2. Click Next.
  3. Review and Accept to the License Agreement. Click Next.
  4. Select an Installation Directory and click Next.
  5. Click Next again to start the installation process.
  6. Click Finish.


Configuration

With a Master Host and Language Pair installed, you must configure the job engines and the related translation engines.

Adding a Job Engine

  1. Go to <MasterHost>:8000
  2. Click Edit in the top-right corner.
  3. Click Manage > Deployment.
  4. Click Add in the Job Engines column.
  5. Enter a Host.
  6. Click Add.
  7. Click Done Editing in the top-right corner.

Adding a Translation Engine

  1. Go to <MasterHost>:8000
  2. Login with the credentials created during installation.
  3. Click Edit in the top-right corner.
  4. Click Manage > Deployment.
  5. Click Add in the Translation Engines column.
  6. Enter a Host.
    • This is the Job Engine address.
  7. Select the Mode as Auto-detect.
    • The language Pair can be selected but will not populate without a License.
    • If you are sure of the language pairs being assigned to a job engine, then provide the language pair details respectively.
  8. Click Add.
  9. Click Done Editing in the top-right corner.

Obtaining a License from RWS

  1. Go to <Hostname>:8000
  2. Login with the credentials created during installation.
  3. Click on Entitlements.
  4. Click Download Entitlements Profile.
    • Myhost.json will be downloaded.
  5. Contact SDLMT Support Team to get a license file.
    • Go to https://gateway.sdl.com
    • Click Login for Support.
    • Enter the assigned SDL Administrator credentials.
    • Click on My Support.
    • Click Log a Case.
    • Within the form provided, provide basic details such as, email addresses, description of the languages required and upload the resultant myhosts.json file.
    • Click Submit Case.
      • Please note, the resultant license file will be sent to you in 1-3 days.

Adding a License

  1. Go to <MasterHost>:8000
  2. Login with the credentials created during installation.
  3. Click on Entitlements.
  4. Click Upload License File.
  5. Locate the license file provided by RWS and click OK.
    • You may be prompt about unassigned language pairs. If you have not assigned a language pair to a job engine, then you will need to do so by navigating to the Hosts page.

Starting/Stopping Job or Translation Engines

  1. Go to <MasterHost>:8000
  2. Login with the credentials created during installation.
  3. Click on Hosts.
  4. Click the Stop/Start button to stop or start an engine.

Verifying Translation Service

  1. Go to <MasterHost>:8000
  2. Login with the credentials created during installation.
  3. Click Translate.
  4. Select an input language and enter some text.
  5. Select an output Language.
    • The results should be displayed. If they are not, please ensure the SDL MTE Manager service is running.

Obtaining the RWS API Key

  1. Go to <MasterHost>:8000
  2. Login with the credentials created during installation.
  3. Click on My Account in the top-right corner.
  4. Copy the API key and save it in a text file for reference.
  5. Copy the API Base URL.


Adding the Translation Server in FTK Plus

  1. Do one of the following:
    • For new cases, in the Add/Import Evidence > Processing Options dialog click Customize.
    • For existing cases, in the Examiner, click Evidence > Additional Analysis.
  2. Check Language Translation.
  3. Click Language Translation Options.
  4. Click Configure.

  1. Enter the following:
    • Server URL
    • Port Number
    • API Key
  2. Click OK.
  3. Select the Types to Process.
    • Documents
    • Presentations
    • Spreadsheets
    • Email
    • Graphics

Adding the Translation Server in FTK Core

  1. Do one of the following:
    • For new cases, in the New Case Options dialog click Customize.
    • For existing cases, in the Examiner, click Evidence > Additional Analysis.
  2. Check Language Translation.
  3. Click Language Translation Options.
  4. Click Configure.

  1. Enter the following:
    • Server URL
    • Port Number
    • API Key
  2. Click OK.
  3. Select the Types to Process.
    • Documents
    • Presentations
    • Spreadsheets
    • Email 
    • Graphics
  4. Select a Target Language.
  5. Click OK.


Multi-server Deployment Example

The diagram below is an example of a hypothetical multi-server deployment with FTK Enterprise, it includes the following:

  • 1 x FTK Enterprise Machine
  • 1 x Worker Host
  • 1 x Job Engine with 2 x Translation Engines.
  • 1 x Job Engine with 1 x Translation Engines.

Hostname

Master Host

Worker Host

CPU

RAM

Storage

Language Pairs Installed

Fiona

Yes

No

4 Core

4GB

60GB

None

Penelope

No

Yes

8 Core

16GB

60GB

French to English

Spanish to English

Hercule

No

Yes

4 Core

8GB

40GB

German to English

 

Host Penelope has been assigned more resources in this example as it will be handling an additional translation engine as well as it being expected to have higher usage.