Moses MT Market Report

This report gives an overview of the landscape of Moses Machine Translation: the history, its adoption by industry, estimates of market size and demographics, types of users and usages, types of players in the market of Moses-backed solutions and their offerings and potential future scenarios for Moses use outside academia. It’s a valuable source of information for everyone who is interested in MT and specifically in the Moses open-source MT system.

Open source machine translation systems do not differ from proprietary systems from a technical perspective. Virtually all approaches to machine translation are available as software code and packages licensed under an open source license, albeit some never developed a large contributor base and their development has slowed down.

The crucial difference of open source development in comparison to proprietary software is that the development of a complex MT system core including a decoder (the software component that performs the translations) is the shared effort of a loosely organized group of expert developers while still allowing the commercial exploitation of the results.

In the field of machine translation, the developers are often academically funded or the original source code comes from a since-abandoned proprietary software development effort. In some cases academic institutions insist on academic licenses which prohibit commercial use. This often creates duplicate development efforts in academia and industry for the same core algorithms and abandoned academic software. In our opinion, allowing the use of academically developed code by anybody for any purpose is justified, as the research is most often publicly funded and it strengthens the development effort for any use, including academic use.

This report explores how the availability of Moses as an open source solution has influenced the MT market, Moses’ contribution to making a broad range of MT solutions available to users and the novel uses Moses enabled in the market as well as in internal use by organizations that would otherwise not have been possible.

Source: TAUS News

NLP News