Linked Data for Language Technology Community Group
This group aims to consult with current and potential users of linguistic data to assemble use cases and requirements for Language Technology Applications that use Linked Data. The results will be used to guide future interoperability, research and development activities spanning the language technology and linked data domains. Potential users are companies and public bodies involved in natural language processing, language resources, content management, the language services and localisation industry and other applications of content analytics techniques used in search, recommender systems, sentiment analysis and terminology management. The group does engage with users through surveys, international events and training activities organized in conjunction with partners from academia or industry, resp. designated research projects and networking efforts (esp., EU or other multi-national projects). We identify use case and requirements priorities, technology gaps and interoperability roadblocks. We work towards community group reports that describe our findings and/or solutions to the challenges identified in our work.
Group's public email, repo and wiki activity over time
Note: Community Groups are proposed and run by the community. Although W3C hosts these conversations, the groups do not necessarily represent the views of the W3C Membership or staff.
In the last 1 1/2 years, the LIDER project has organized several roadmapping events to gather a broad community around the topic of linguistic linked data. In July this year, LIDER will engage with two selected communities. On July 6, the 5th LIDER roadmapping workshop will be held in Rome at Sapienza University of Rome. The topic will be cross-media linked data and the event will provide several high level speakers from the multimedia area. On July 13th LIDER will organize the 6th roadmappping workshop in Munich. The event will be hosted by Siemens and will focus on content analytics and linked data in healthcare and medicine.
For both workshops participation is limited. If you are interested in the Rome event please contact Tiziano Flati, for Munich please contact Philipp Cimiano.
Following successful road-mapping workshops at the European Data Forum and the Multilingual Web conference, the LD4LT group will next be engaging with the Localisation Industry on the 4th June’14 in Dublin, Ireland. This will be part of the FEISGILTT workshop on harmonising localisation industry standards, which is co-located in Localisation World, the industry’s premiere trade show. The focus will be on use cases where linked data can contribute to interoperability for localisation tools and processes, including terminology management and parallel text curation.
The use of Linked Data for Language Resources is a hot topic at the Language Resources and Evaluation Conference (LREC’14), 26-31 May, Reykjavik, Iceland. The week kicked of with a LD4LT Tutorial. All LREC delegates are able to contribute by completing a questionnaire included in the conference bag (and enter a prize draw) and can talk to existing LD4LT members at the LIDER/FALCON booth in the EU Village.
The results of an initial survey conducted by the LIDER project into requirements and use cases related to linguistic linked data are now available. It is interesting to read this in tandem with results of a survey just released by LT-Innovate on interest in a European Language Cloud.
The Linked Data for Language Technology community is organising a roadmapping workshop on 21st March in Athens, to build a better understanding of the potential synergies and co-evolution paths for language technologies, such as machine translation, information extraction and sentiment analysis, and linked data. Language technologies are key to extracting information from unstructured content in different languages to form linked data, while linked data can aid the discovery and sharing of the language resources that underpin language technologies.
Who should attend? Any organisation interested in automated extraction of data from unstructured digital content, especially content in more than one language and including multimedia as well as textual content. Organisations engaged in the market for language technologies applied beyond English-language content and data. All these can benefit from more open access to linked language resources.
How can you participate? You can register for the event here. If you wish to present a similar statement you can indicate this on your registration form. The event will then proceed in an structured open format to identify and capture from participants their use case priorities and interoperability, best-practice and technology gaps they face. An online survey is currently open for gathering industry view on use case prioritation. You can also contribute directly by joining the Linked Data for Language Technology community at the w3C
Programme and Topics: The workshop will open with keynotes from Hans Uszkoreit who is Scientific Director DFKI, Nicoletta Calzolari Director of Research CNR, Phil Archer who is leading the W3C Data Activity and Asun Gomez-Perez UPM who is leading the LIDER coordination action on linguistic linked data. This will be followed by short briefing from four existing international communities working in this area, by position statements from companies about existing use cases and by an open workshop session to establish use case priorities.
The language resource community has already made a concerted attempt to catalogue different data sets through the META-SHARE initiative. It has tackled the need for common meta-data for linguistic corpora of various types and has paid particular attention to encoding the different usage rights that exist across governmental, academic and commercial data sources. This initiative is therefore well primed to exploit linked data technologies being standardised by the W3C Data Activity to further open the cataloguing and discovery of language resources.
This is particularly timely as the European Commission has launched it new H2020 funding programme with a strong support available for innovation and research in the open data and language technology space. In April 2014 it will also launch its Connecting Europe Facilities programme, with €1Billion for funding new pan-European digital services, including open data exchange and automated translations services. In both these initiatives, strong, open solutions for the interoperability of language resources as open web data will be key.
The workshop we take a use case driven approach to key questions around the synergies possible between the W3C’s open web data standards and existing approaches to sharing language resources and applying them for training language technologies:
How can language resource sharing infrastructure, such as META-SHARE, migrate to a linked data approach so as to benefit from more robust, decentralised and scalable publication and search features?
How can emerging onto-lexical resources such as BabelNet be usefully interlinked with individual terms in existing language resources?
How can the process of locating and managing language resources to train language technologies be eased and optimised by vocabularies such as the Provenance Ontology or the Provenance and Plans Ontology for repeatable data workflows.
However these are just a sample of the many issues and viewpoints that will have a bearing on the future of Linked Data for Language Technoloiges, and we hope you will be able to join us in Athen to share yours.
Organisations world-wide are struggling to better use the WWW to engage in meaningful online conversations with customers and citizens. To do this in a scalable and cost effective way, many are turning to automated language technologies. These can assist in: discovering/extracting information; understanding opinions/trends; processing and managing multilingual/multimedia content and data; and monitoring/forcasting topics of interest.
However, if you are already considering or using language technologies you will understand the key role played by data in training automated language technologies to meet the needs of your specific application. Locating, collecting and determining the quality and relevance of such linguistic data therefore forms a major cost, and a barrier, for the successful use of language technologies.
Open linked data on the web, using standards developed by the W3C, may offer an ideal solution to discover and exchange linguistic data across a wide range of commercial and governmental applications. However, establishing international best-practice and developing open technical specifications requires a much better understanding of these different applications and their requirements.
To this end, a new W3C community group has been formed to assemble and discuss use cases and data handling requirements for language technology applications. We invite you now to join the Linked Data for Language Technology (LD4LT) group and engage in these activities. You can provide an indication of your particular interests and requirements via the initial survey at https://www.w3.org/2002/09/wbs/1/ld4lt-wbs1/
We also invite you to participate in any of the upcoming road-mapping workshops being organised by the group at the following events:
The results of these consultations will be published for discussion via the LD4LT group. They will provide a roadmap for other interoperability, research and platform development activities spanning the language technology and linked data domains. These activities will include the W3C OntoLex and Best Practice in Multilingual Linked Open Data community groups as well as future EU-funded collaborations under the H2020 programme.
ld4lt
