Embed Size (px)
DESCRIPTION
This is an update regarding the EU-funded FALCON project which provides a web-based platform integrating commercial tools Easyling, XTM Cloud and TermWeb with Moses core-based SMT from Dublin City University and the use of linguistic linked data in localization workflows through a schema provided by Trinity College Dublin. The platform is compliant with Linport and other currently promoted standards.
Citation preview
Wednesday, 4 June /09:40 – 10:10
Ac5ve Terminology Promp5ng for SEO and Website Transla5on
Ioannis Iakovidis, Interverbum Technology
TaaS Workshop 2014 4 June, Dublin (Ireland)
The research within the project TaaS leading to these results has received funding from the European Union Seventh Framework Programme (FP7/2007-‐2013), grant agreement no 296312
Welcome Ivan Smolniov, ABBYY Language Services
Federated Active Linguistic data CuratiON
Ioannis Iakovidis, MD, Interverbum Technology
FALCON
In brief… § Interverbum Technology est. 1999
§ Today more than 80,000 licensed TermWeb users
§ Some clients:
Bauer, Euroscript, Arancho, FujiXerox, Medtronic, Novell, SAS Institute, Scania, SDI Media, SIS (Swedish Standards Institute), Xerox, VMware, Salesforce.com, Agilent, The World Bank
§ Clients served from five offices in Europe, North America and Asia
Terminology Management for Global Marketing
Web-based Open architecture
Easy administration
Visuals and sounds
TermWeb 3.11 The most powerful terminology management system anywhere
COLLABORATE quick and easy grid-view editing
AUTOMATE
powerful customizable workflows
INTEGRATE cross-system communication by TermWebIntegrator
EMBELLISH
full media support
EVOLVE platform-independent, cloud-based, uniquely versatile
TermWeb 3.11 Brings easier collaboration to the cloud
FALCON in a nutshell
§ L10n is a Big Data Industry
§ Large-scale, Monetised Reuse of Translation Memories and Term Bases
§ SMT and Text Analytics now also leverage these resources
§ Large clients and LSPs curate and add value to such resources as assets
§ FALCON aims to extend these benefits to SMEs
Leverage the power of Linked Data for the Long Tail of the Localisation Industry
Multilingual Web
Language Resource
Community
Language Technology Community
Localisation Industry
XLIFF1.2
ITS 1.0
ITS 2.0
XLIFF 2.0
MLW-LOD
NIF-RDF
FALCON
§ Trinity College Dublin (IE) LOD Mapping and Link Quality Federated Access Control L10n Interoperability
§ XTM International (UK) CAT/L10n management vendor and interoperability
§ Dublin City University (IE) SMT and text analytics
§ SKAWA Innovation (HU) Web site translation (EasyLing), crowdsourcing
§ Interverbum Technology (SE) Terminology Management
FALCON Consortium
§ Provide an Open Schema and SaaS platform for exposing language resources as linked data
§ Enable controlled, decentralised sharing of resources and stand-off value-add annotation
Term or named entity annotation Translation process provenance and QA
§ Active Curation of resources and value add
§ End-to-end process management
§ On-demand assembly of domain specific LT training corpora
FALCON Approach
From: http://www.w3.org/TR/prov-primer/
ITS related entity subclass: document, segment, analysed-text, term, translation, translation-revision
subproperty: wasTranslatedFrom
W3C Provenance WGhttp://www.w3.org/2011/prov/
Provenance-Oriented Approach
Language Resource LOD Store
Public Language LOD
Resource Curator
Multilingual Web Management (EasyLing)
Translation Management (XTM Cloud)
Machine Translation
(Moses – DCU)
Terminology Management (TermWeb)
Client CMS
Text Analytics (NER – DCU)
L3Data API L3Data API L3Data API L3Data API
Use
rs
Sys
tem
s Li
nked
Dat
a
Localisation Client
Project Manager
Translators/ Posteditors
Translation Reviewers
Terminologist
Source doc
Target doc
Source seg Source
seg Source seg
Target seg Target
seg Target seg
bi-text bi-text bi-text
bi-text bi-text ML terms
Project TM
Project term base
QA meta-data QA meta-
data QA meta-data
TB
TB
TB
TB
TB
CMS/SEO CMS/SEO TermWeb
Local termbases with SEO-scores are linked via FALCON and translated in TermWeb
TB
FALCON
TB
TB TB
TB
TB
TB
Local termbases with SEO-scores are linked via FALCON and translated in TermWeb
The translated term with the best SEO-score will be returned to the local termbase
FALCON CMS/SEO CMS/SEO TermWeb
§ Language Resource Publishers can audit links used in building other resources, track ROI
§ Tool Vendors and Integrators expand markets with more open asset management offerings
§ SME LSPs gain resource sharing and pooling opportunities that avoid lock-in
§ LSPs and clients can use Active Curation to quickly train domain specific SMT and text analytics components
Value Network of Benefits
Welcome to TermWeb
§ Interverbumtech.com
§ +46 13 32 98 40 (or any other office)
STOCKHOLM LINKÖPING CHICAGO BOISE SINGAPORE
