GTS Blog

The first day of the Multilingualweb workshop in Pisa is over. About 100 people attended day 1. It was an extremely busy session with about 20 different talks. Overall the organization of the workshop was good. Speakers were limited to 15 minute timeslots and the chairs made sure that everyone completed their talks on time.

Kimmo Rossi of the EU discussed some of the funding opportunities available to European language technology companies. A case in point is the SME Initiative on Digital Content and Languages which has allocated 35 Million Euro for projects. That call is ending in about 3 weeks. Mr. Rossi’s group will not be initiating funding calls for another year or so.

Since the conference was organized by the World Wide Web Consortium (W3C), a lot of the talks focused on web standards. Personally, that kind of stuff bores the heck out of me. Very technical and I just don’t see the point. So pardon me if I don’t spend too much time writing about it lest I fall asleep at the keyboard. Italy’s W3C chief Dr. Oreste Signore did, however, manage to put together a lively and interesting talk about web localization standards.

Ralf Steinberger who heads up the European Commission’s JRC (Joint Research Centre) gave an overview of his group’s activity. Which you can see online here. This set of web services is being used by about 2 million people a day for news and data retrieval in a few dozen languages. Dr. Steinberger, who started as a one man show in 1998 has grown a moderately large team which also includes a group of about 7 PhD level computational linguists. They are doing some pretty high level stuff like sentiment analysis and named entity extraction. They are also developing their own machine translation capabilities around the Moses open source system. They are currently licensing MT from Language Weaver in two language pairs. But decided to develop their own due to the high cost of MT software licensing.

Jochen Leidner of Thompson-Reuters gave an interesting talk on how his company maintains a $13 Billion a year business by providing high quality content to its customers, navigating through the diverse data sources, standards and markup technologies. Dr. Leidner recently relocated from the US to Zug, Switzerland to head up a new R&D group.

Sophie Hurst of SDL reviewed how her company is helping their clients get their message across to markets all over the world quickly and efficiently. SDL provides web globalization services to some of the biggest companies in the world.

Pal Nes of Opera Software, a Norwegian software company that develops web browsers, discussed how his company used crowdsourcing to localize their software into various languages. Opera used Pootle, an open-source platform which I wrote about in a previous blog post.

There were other very good talks but I won’t discuss every one of them. The slides for most/all of the talks are available on the Internet and you can read more about it there.

One thought in summary: the biggest players in the Internet and software industries are US companies that never received any government funding. They came up with winning products which won over users worldwide. It’s called the ‘American way.’ Innovation, hard work, no bullshit. No free lunches of any kind. The multilingual web has tremendous potential, but if Europe wants a leadership position in it they will have to beat Google, Facebook and a bunch of other US companies. I say: good luck!