Mstering the language challenge with the public sector and a Google tool
Balancing the project language is a challenge in international projects or in projects when experts from one country work in another country. Although the English language is a very good common denominator mastered by most of us there are a lot of documents which are not translated – e.g. country-specific documents with legal requirements or simply a mail from an end-user with important information for a blueprint which needs to be read very quickly.
A tool from Google: Google translator has helped us a lot over the last months in several projects. By using this tool we got a good overview over several documents and mails in foreign languages like Danish. There are some limitations for sure: You cannot just write a blueprint in German and then just translate it via this tool to Danish. A lot of follow-up work needs to be done then. But it works much better the other way round: Just try it out and copy a mail in a foreign language, past it to Google translator and check the result.
When I noticed the new Google application for the first time I was quite sceptic as I had disappointing experiences with machine translation tools during my studies a long time ago. But after the first promising experiences I dig a bit deeper and found out the foundation for the improvements: Google uses a large amount of documents which are available in several languages, especially documents from the United Nations and the European Union. Due to the fact that these organizations have several official languages (the UN has 6) the statistical machine translation tools have an ideal starting point as they could easily match the words and phrases together in the different language versions.
A feature which I use very frequently is the possibility to propose a better translation in order to enhance the translations provided by the tool in the future. Although a lot of translations needed for documents used in SAP projects are already available some translations are misleading. Therefore together with some colleagues which are also using the tools quite frequently I have proposed better translations. In many cases we have already noticed an improvement for the phrases. For me this could be really a community approach: Let us make this tool even better for the use by SDNer and BPXer.