SBIR Phase II: Software-as-a-Service Customized Machine Translation for Commercial Language Service Providers and Their Clients

这个小型企业创新研究第二阶段项目开发了先进的技术能力，用于在商业环境中构建和部署客户端自适应的自动化语言翻译系统，这些系统由全球化企业和为此类企业客户提供翻译服务的语言服务提供商公司（LSP）使用。 开发的技术利用先前翻译材料的数据库，为商业语言服务提供商（LSP）及其企业客户提供客户适应的高质量全自动翻译。这种方法为创建和部署客户特定的定制机器翻译（MT）引擎提供了可扩展且成本较低的解决方案。 一旦部署，这些定制的机器翻译系统将扩大客户的翻译能力，使其能够翻译使用当前方法无法翻译的大量内容。该项目更广泛的影响/商业潜力在于它将对广泛的商业翻译行业产生影响。 该项目开发的技术可能会大大减少障碍，广泛采用MT技术的广泛LSP行业和他们的企业客户群。2010年的商业翻译市场是一个价值260亿美元的行业，并以健康的速度增长。目前的商业MT产品价格昂贵，对于大多数企业和服务提供商来说太难部署。 基于网络的免费翻译服务为临时用户提供服务，但不能满足企业的质量和安全需求。该项目开发的技术和基于云的交付模型支持可扩展、易于集成的机器翻译服务，这对广泛的潜在客户具有高度吸引力。 这种方法将支持以多种目标语言大规模地生成具有成本效益的内容，这种能力对于全球化的美国企业来说是必不可少的，以便在世纪信息丰富的市场中竞争。

This Small Business Innovation Research Phase-II project develops advancedtechnology capabilities for constructing and deploying client-adaptedautomated language translation systems within commercial settings that areused by globalizing enterprises and the language service provider companies(LSPs) that provide translation services to such enterprise clients. Thedeveloped technology leverages databases of previously-translated material inorder to produce client-adapted high-quality fully-automatic translations forcommercial language service providers (LSPs) and their enterprise clients.This approach provides a scalable and less-costly solution forcreating and deploying client-specific customized Machine Translation (MT)engines. Once deployed, these customized MT systems expand the capabilitiesof clients to translate volumes of content that are not feasible to translateusing current methods.The broader impact/commercial potential of this project lies in the impactthat it will have on the broad commercial translation industry. Thetechnology developed in the project is likely to significantly reduce barriersto wide-spread adoption of MT technology by the broad LSP industry and theirenterprise client-base. The 2010 commercial translation market is a $26billion industry, growing at a healthy pace. Current commercial MT offeringsare expensive and too difficult to deploy for most enterprises and serviceproviders. Free web-based translation services serve casual users, but do notmeet the quality and security needs of enterprises. The technology developedin this project and the cloud-based delivery model support scalable,easy-to-integrate MT services, which are highly attractive to a broad range ofpotential clients. This approach will support cost-effective content generation intomultiple target languages at a massive scale, a capability that is essentialfor globalizing US enterprises in order to compete in the information-richmarket place of the 21st century.