Semantic interoperability for the multilingual web

2367

“Semantic interoperability for the multilingual web” in the context of the Public Multilingual Knowledge management Infrastructure for the digital single market (PMKI) project was the name and the theme of a two-day workshop organised and hosted by the Publications Office of the European Union on 4-5 June 2019.

TermCoord attended it as member of the PMKI Steering Committee and as terminology representative. The workshop provided a platform for presenting two closely related ISA2 actions (PMKI and VocBench) whose development is led by the Publications Office.

The PMKI is an amazing initiative intended to collect language resources (vocabularies, thesauri) of the EU and Member States and to make them interoperable. It aims to create a public multilingual knowledge infrastructure that supports companies, in particular in the language technology industry, with the implementation of multilingual tools and features needed to improve the cross-border accessibility of e-commerce solutions.

The PMKI will provide a platform based on VocBench, an already existing web-based open source multilingual collaborative tool for managing OWL ontologies, SKOS (XL) thesauri, generic RDF datasets, as well as Ontolex-Lemon resources. In a nutshell, the workshop was a two-day encounter of semantics, terminology and technology during which the Publications Office teams behind the PMKI, VocBench and EuroVoc presented the results of both ISA² actions with live demos of these tools in their current states of development and implementation. Additionally, they presented ideas and related projects. External participating experts provided a very positive feedback and appreciation.

Several speakers were invited as leading expert in the semantic web domain to present and to share their experiences in the area. Participants were from different communities: scientific community, public administrations, Language Technology Industry, colleagues from different EU institutions.

Dr John McCrae was there to present OntoLex-Lemon, Dr Armando Stellato presented the next version of VocBench and Roland Wingerter from Yellow Pages was there to present the thesaurus and its integration to VocBench as a case of best practice of private enterprises which engage with the PMKI project.

Jochen Hummel from Coreon presented the Neutral Machine Translation workflows with the subject matter expert as the key human factor involved in the (automated) translation process and the linguist (translator, terminologist, etc.) as a multilingual knowledge manager.

A speech we had already heard at JIAMCATT 2019 and which also this time made quite an impression on the people present but to a lesser extent than at JIAMCATT which is due to the more technological approach of this workshop and thus profile of the attendants.

Dr Tom Baker presented the conceptual mapping and experiences from GACS and as a result of Roland Wingerter’s input and a question about automating the process, a spontaneous presentation was arranged and given by Dr Cédric Pruski and Dr M. Da Silvia from Luxembourg Institute of Science and Technology about their research on exactly that: automation of the maintenance of conceptual mapping. Please find all the presentations below this article.

Each of the two workshop days had discussion rounds included and due to the many times academic and technical profile of the speakers, the workshop took a rather academic shape which allowed us to get an insight into the newest research and technologies within the areas of semantics and interoperability. You can consult the full program of the workshop here.

All in all a very interesting and new approach which goes hand in hand with our new investment in machine translation. TermCoord is happy to be able to actively contribute to PMKI.

Read the more about the PMKI project, VocBench, EuroVoc and this event here.

You can also find all the presentations here.


Written by Stine Jensen, Terminologist at TermCoord