FREME: Coupling Language and Knowledge via e-Services Ecosystem

FREME is a 2-years innovation action which aims at creating a commercial-grade open framework of e-Services for multilingual and semantic enrichment of digital content. FREME validates its technologies in 4 real life business cases: digital publishing, translation and localisation, agriculture and food domain, and personalisation.

In the last few years, the growing amount of digital content across languages, sectors and borders led to both challenges and business opportunities for many industries. Linked Data (LD) and Language Technology (LT) solutions exist and have gained popularity, providing useful technologies such as machine translation, entity recognition, or multilingual linked data sets. Nevertheless, their coupling in real-world business is limited due to several issues like: a plethora of content formats to process, adaptability and “silo solution” dependency, and usability in an industry application scenario.

FREME – a two-years market-oriented H2020 Innovation Action funded by the European Commission that started in February 2015 – will make a step forward to overcome these issues. FREME will address the challenges to validate that multilingual and semantic technologies are ready for their integration in real life business cases. On the one hand, the project brings together top-tier specialists in language resources, tools, and services, multilingualism and linked data and open data for managing technical aspects. On the other hand, the design of core technological features in FREME is driven by four different business partners that are integrating FREME in their business solutions.

The figure shows the whole concept and approach of FREME. Via the framework, various user types (e.g. interface developers) can combine the e-Services easily to get additional value:

• e-Translation, to access domain adaptable machine translation;
• e-Terminology, to access terminology identification and linking to term data bases;
• e-Entity, to access identification of named entity recognition with user defined data sets for entity linking;
• e-Link, to access integration of additional information from linked data sources;
• e-Internationalization, for processing formats like HTML, XLIFF or general XML; and
• e-Publishing, for storing enriched HTML as eBook content.

The development of the FREME open framework has been carried out taking care of four main requirements among many others: flexibility, scalability, robustness, and usability. Scalability is assured by the code that has been structured to ease an extension of the framework; with the Alpha version target users are able to easily create their own services and to benefit from a combination in the forehand described manner. Robustness of the framework has been considered by continuous deployment of the relevant software development methodologies like unit testing and or performance monitoring. Flexibility and usability have been covered by putting several standards in the realm of multilingual content into action, especially those from the area of so-called linguistic linked data.

FREME is providing a standards-based open framework of e-Services that already been released in its Alpha version, available for using and testing by third parties. The framework provides access to the six e-Services via a reusable set of open Application Programme Interfaces (APIs) and Graphical User Interfaces (GUIs). The framework will improve and extend existing processes related to management of digital content, going through the whole content value chain: content creation (or authoring), content translation and localization, publishing and access to content including cross-language sharing and personalized content recommendations.

Source: http://cordis.europa.eu/news/rcn/132754_en.html?isPermaLink=true?WT.mc_i...