FREME: Multilingual Semantic Enrichment with Linked Data and Language Technologies

FREME: Multilingual Semantic Enrichment with Linked Data and Language Technologies 1,2Milan Dojchinovski, 3Felix Sasaki, 4Tatjana Gornostaja, 1Sebastian Hellmann, 5Erik Mannens,5Frank Salliau, 6Michele Osella, 7Phil Ritchie, 8Giannis Stoitsis, 9Kevin Koidl,1Markus Ackermann,1Nilesh Chakraborty 1InfAI, Germany; 2FIT CTU in Prague; 3DFKI, Germany; 4Tilde, Latvia; 5iMinds, Belgium; 6ISMB, Italy; 7Vistatec, Ireland; 8Agro-Know, Greece; 9Wripl, Ireland 1surname@informatik.uni-leipzig.de; 2milan.dojchinovski@fit.cvut.cz; 3felix.sasaki@dfki.de; 4tatjana.gornostaja@tilde.lv; 5name.surname@ugent.be; 6osella@ismb.it; 7philr@vistatec.ie; 8stoitsis@agroknow.gr; 9kevin@wripl.com Abstract In the recent years, Linked Data and Language Technology solutions gained popularity. Nevertheless, their coupling in real-world business is limited due to several issues. Existing products and services are developed for a particular domain, can be used only in combination with already integrated datasets or their language coverage is limited. In this paper, we present an innovative solution FREME - an open framework of e-Services for multilingual and semantic enrichment of digital content. The framework integrates six interoperable e-Services. We describe the core features of each e-Service and illustrate their usage in the context of four business cases: i) authoring and publishing; ii) translation and localisation; iii) cross-lingual access to data; and iv) personalised Web content recommendations. Business cases drive the design and development of the framework. Keywords: Linked Data, NLP, multilinguality, enrichment 1. Introduction In the last few years Linked Data and Language Technolo- gies have reached a state of maturity. From only 294 Linked Datasets in September 2011, the Linked Open Data (LOD) cloud has grown to 1,091 datasets in April 2014 (Schmacht- enberg et al., 2014). These developments triggered number of research in assuring certain level of quality making the data ready for consumption. Language Technologies, such as machine translation systems, entity recognition and termi- nology systems have also reached a certain level of maturity. Nevertheless, there is still a gap between the actual busi- ness needs and the language and linked data technologies, which needs to be bridged; usage domains usually differ and the tools need to be appropriately adopted, multilin- guality and language coverage is not well supported, the tools are offered as ad-hoc solutions without standard in- terfaces, and existing solutions for enrichment of content is limited only to specific knowledge bases. In this paper, we present FREME, an open framework with e-Services for multilingual and semantic enrichment of digital content, which aims at developing APIs and GUIs in order to fill this gap and offer industry ready solution. 2. FREME e-Services The FREME framework offers a set of e-Services for mul- tilingual and semantic enrichment of content. The services are available as RESTful APIs. The design of the services was driven by the requirement of having interoperable APIs by using the NLP Interchange Format (NIF) (Hellmann et al., 2013) format and the Internationalization Tag Set (ITS) Version 2.01. The framework defines following set of e-Services: 1http://www.w3.org/TR/its20/ • e-Entity: perform named entity recognition (NER), classification and linking of entities in multilingual texts. The e-Entity is independent of the back end en- gine and currently it can be utilized with FREME NER and the DBpedia Spotlight2 NER engines. While DB- pedia Spotlight performs entity …

Content truncated to 500 words