Product Information
Transliteration/Translation of Master Data from Non-Latin to Latin Scripts and Vice-versa
About:
This blog gives you a short overview of the implementation of QTC (Quanto Transliteration Center) in SAP to convert Master Data from non-Latin to Latin Scripts with machine translation.
Quanto Transliteration Center (QTC) is a solution devised for transliteration from non-Latin languages like Chinese, Japanese, Korean, Russian etc. into languages with Latin characters, like English.
Initial Situation:
In order to comply with legal requirements or to follow well-established business practices and corporate policies, many multinational companies dealing with countries with non-Latin writing systems have to create their business partners with addresses in the local script, along with an international Latin (English) version which enables master data to be globally readable.
The manual input of these data is very time-consuming. Moreover, both versions must always be up-to-date and consistent, so if you need to change the original script (the local script), you have also to adjust the Latin script or vice versa. Over time, this has led to a number of customer enquiries, requesting a solution to streamline the translation of address versions and obtain master data in both local and international version.
Access:
QTC runs on the SAP BTP and you can access it via RESTful API or using the UI. QTC is easy to integrate via OpenAPI Specification into your existing system landscape. Predefined data models for SAP R/3 and SAP S/4 systems are available.
Content:
There are different scenarios where clients might need translation/romanization of master data:
- SAP S/4 HANA, SAP R/3 Business Partner Maintenance
- SAP Ariba, SAP HCM, SAP CRM
- Any other IT system
- Without direct integration, via QTC frontend
Example:
In this example we want to show how to convert a Business Partner from Chinese to Latin script in SAP/4 HANA:
First of all you have to choose in the Customizing settings whether you want to create a new BP or change an existing one in the international version. If you confirm these settings, QTC is activated in the background and the international version is created.
Then you call transaction BP to display your Business Partner
The international address is initally empty.
By clicking on Save, QTC is activated and both the Chinese and the machine translated international version of your BP are generated.
In this way you can integrate machine translation in the process of creating or changing your business partners.
With this process you can get machine translation of a single BP or also of bulk master data, which of course you can also manually change, if required.
Other types of master data like material master data can be converted as well.
Conclusion:
We have described how machine translation for business partners in SAP/4 HANA or SAP ERP works. QTC is available on the SAP Store: https://store.sap.com/dcp/en/product/display-0000059797_live_v1/transliteration-center-qtc
Your feedback is welcome, feel free to enter your suggestions.
Hello Isabella
thank you very much for sharing! To my curiosity - is there any comparison available of QTC quiality of transliteration with ICU library whcih is available on Netweaver-based SAP solutions?
SAP note : https://launchpad.support.sap.com/#/notes/1062237
ICU documentation: http://unicode.org/reports/tr35/#Unicode_Language_and_Locale_Identifiers
Thank you!
Vlad
Hi Vlad,
thank you for your feedback.
Our solution does not „simply“ transliterate from one script to another. Especially for company names, you need a smart mix of transliteration, translation and data mining.
For example, if we just transliterate the Chinese company name 資誠聯合會計師事務所 , the result would be:
Zicheng lianhe kuaijishi shiwu suo.
Transliteration + translation of this company name would be:
Zicheng United Accounting Firm.
This result is better than the previous one, but still it is not completely correct.
The correct name of this company is:
Price Waterhouse Coopers, a famous company which provides legal and
accounting services all over the world.
So you see how complex and challenging the conversion into Latin script can be and why a flat transliteration does
not always fit the purpose. With our tool we are offering an intelligent mix of transliteration, translation and data mining.
Best regards,
Isabella
Hi Isabella ,
Thank you for clarifying. You are right - when I finished reading the blog, I got impression that it's a classical tranliteration similar to what ICU library is doing. Now with your explanations it's more clear and makes a good sense
Thank you, very nice!
Vlad