Sharing collective knowledge is fundamental to better decision making and problem resolution. In the age of the Internet, Unicode and on line file systems, sharing data is easy, but a difficult barrier remains which is language comprehension. Often knowledge management systems start out to support a single language and as the system grows the requirements for multilingual support evolve. In these scenarios human translation is often prohibitive because of the volume, cost and dynamic nature of the collective data. Under certain circumstances Machine Translation can be an ideal solution.

This presentation will talk about Oracle's experience in implementing a pilot project to add Machine Translation to an online support system that previously only supported English. We will cover what Machine Translation is, what is customization and why it is essential, what can be achieved, and go through the process of implementing a system leveraging a 3rd party Machine Translation product. We will conclude this presentation with a live demonstration of this online support system, in which we will do some cross- language searching and on the fly translations to show the discussed concepts actually implemented.