Published in Budapest Business Journal, September 2012
The relationship of IT and translation has always been close, language being at the heart of both, and continues to evolve. We can see how the unabated exchange of ideas and concepts has been shaping them, especially in the last decade. The IT boom not only brought along new tools for the language experts to play with, but also fundamental changes in approach, and gave rise to completely new markets, previously either non-existent or out-of-reach. Computer-aided translation environments facilitated skyrocketing productivity and processing speeds, in addition, automation and smart quality assurance practices have become widespread.
However, this is not a one-way path. As the localization industry has matured, the IT sector started to employ the techniques of internationalization as a means to expand the consumer base. Developers use localization-aware tools, and build products which do not have to be reverse-engineered to comply with any locale, and certain interoperability formats and encoding systems have grown into de facto standards. IT companies thrive on new-fangled, innovative solutions, and they are keen on implementing ideas coming from localization. Some, such as crowdsourcing have yet to live up to the expectations, as the case of Facebook clearly showed, but others have gained acknowledgement. Machine translation, a branch of AI research, has been progressing as a joint effort, and appears to be the new big thing. While this impression may be misleading, because MT is one of the oldest tricks in the sci-fi book of dreams, recent advances and increasing computation speeds allow MT to be applied successfully.
On the borderlands between IT and translation other cutting-edge technologies and ideas have emerged, including data mining, exploration of the deep web and real-time business intelligence solutions. On top of the conservative b2b and b2c targets, such technologies will eventually render c2b and c2c content fit for globalization. These directions would not have been feasible just a couple of years ago, but mining unstructured, yet interconnected corpora will prove not only fascinating, but also relevant to business.
For those of you with an inclination for the mundane technical aspects, let me give a quick peek into the rabbit hole of IT localization without submerging into the dreary details.
When it comes down to practicalities, localization can offer solutions to certain distinctive features of IT content while taking the sensitive time-to-market factor into consideration. There are established methods to manage non-linear text with high-frequency updates, while ensuring consistency and version control. In the case of non-linear text, such as the strings of a user interface, context is essential to produce proper output, and flying blind usually takes its toll on quality. A close cooperation between the developer and the localization team can make a considerable difference. There are certain measures which can prevent scope creep and discrepancies, especially when constant updates are a factor.
- Clearly defined project paths and pause points streamline management;
- A comprehensive term base is the best way to ensure consistency;
- A secure, server-based, online collaborative translation approach improves productivity without compromising quality;
- And last but not least, applying meta-data on all resources eliminates versioning issues and incorrect reuse of leveraged translations and therefore also reduce costs.
The future of IT and localization may not be shaped as we would imagine it years ago it years ago, with silicon-based intelligence, universal translators and the Babel fish. But localization 2.0 is now on our doorstep, and we do not have to wait long to see these new technologies realized and introduced into everyday production.
Originally published in Budapest Business Journal, September 2012