Finished projects

During the past few years, the Dutch Language Institute (formerly Institute for Dutch Lexicology, INL) has worked together with other parties on various national and European projects. A selection is presented below.

Letters as Loot

• Time span: 2008-2013
• Links: www.brievenalsbuit.nl, Sea letters setting course for readability (news article)

In the Letters as Loot project, 17^th– and 18^th-century letters between Dutch people far away from home (such as sailors) and the home front were linguistically researched at Leiden University. The letters have been manually transcribed (typed out) by volunteers from the Leiden-based project Wikiscripta Neerlandica. The collection of letters is a goldmine for historians and linguists, because little is known about the everyday language of that period. The INL contributed to the linguistic annotation of the letters. That means that all kinds of linguistic information, such as spelling and word class, was added to the transcribed words. The addition of a modern standard form to all words and forms (lemmatization) makes its easier to search the texts. The INL has also developed a search application that was launched on 5 September 2013.

CLARIN

• Acronym for: Common Language Resources and Technology Infrastructure
• Time span: 2009-2013
• Links: overview of all sub-projects, clarin.eu

CLARIN’s objective is to realize an integrated, interoperable research infrastructure using language sources and language technology. The infrastructure, which is to be stable, permanent, accessible and expandable, must put an end to the current fragmentation and promote the use of computational technology in the humanities (eHumanities). Within CLARIN, the INL was involved in several sub-projects, such as Metadata, TICCLops, DUELME-LMF, GTB-WFT, VU-DNC and NAMESCAPE.

DAM-LR

• Acronym for: Distributed Access Management for Language Resources
• Time span: 2005-2006
• Website: www.mpi.nl

In the DAM-LR project, an infrastructure was developed providing access to and management of language archives. Examples of language archives include (multimedial) corpora, lexica, grammars, etc. DAM-LR was a small-scale version of European CLARIN. Several parts of the (functional, prototypical) infrastructure of DAM-LR have been further developed in CLARIN.

ELAN

• Acronym for: European Language Activity Network
• Time span: 1998
• Website: cordis.europa.eu

The objective of ELAN was to make all data from the European research community uniformly accessible, using a single query language and a single user interface. In order to make the data from specifically PAROLE and TELRI more accessible, the Dutch Language Institute, together with the University of Birmingham and Università di Pisa, has defined a Common Query Language.

ENABLER

• Acronym for: European National Activities for Basic Language Resources
• Time span: 2001-2003
• Website: www.ilc.cnr.it

ENABLER was an infrastructural project with the objective of promoting cooperation between national activities in the field of ‘Language Resources’ (such as corpora and lexica). The Dutch-speaking regions were represented by the Dutch Language Union and the Dutch Language Institute, among other institutions.

ENeL

• Acronym for: European Network of e-Lexicography
• Time span: 2014-2018
• Links: COST-website, dictionary portal for European languages

The aim of the project was to set up a European network for lexicographers that increased the visibility, accessibility and use of European national dictionaries. To that end the network regularly organized international meetings, workshops and training courses. Together with partners from 28 other European countries, the INL and the Fryske Akademy were responsible for the project application with COST: an organization that stimulates scientific cooperation in Europe. Within the ENeL project the INL functioned as Grant Holder and was part of the steering committee.

IMPACT

• Acronym for: IMProving ACcess to Text
• Time span: 2008-2012
• Website: digitisation.eu

IMPACT was a European project that made online historical texts more accessible. In order to achieve this, the quality of the digitization and search methods in historical texts were optimized. Read more.

PAROLE

• Acronym for: Preparatory Action for Linguistic Resources Organisation for Language Engineering
• Time span: 1991-1999
• Website: cordis.europa.eu (PAROLE), cordis.europa.eu (LE-PAROLE)

The PAROLE project catered to a growing need for modern electronic language sources, to be used mainly for language technology. Standards for the construction and enrichment of text corpora and text lexica were formulated on the basis of evaluative and feasibility studies. Institutes from 14 European countries cooperated on this project. Subsequently, a text corpus and a computational lexicon were built for each of the 14 corresponding languages, following the same standards. As a representative of the Dutch-speaking world, the INL made the following products:

• PAROLE Corpus
• PAROLE Distributable Corpus
• PAROLE Lexicon

SIMPLE

• Acronym for: Semantic Information for Multifunctional Plurilingual Lexica
• Time span: 1998-2000
• Website: clul.ulisboa.pt

The objective of SIMPLE was to add semantic information to the morphological and syntactic information already in the PAROLE lexicon. The information was selected on the basis of its usefulness to language technology applications. The INL served as the representative for the Dutch-speaking regions in this project.

Spelspiek

• Time span: 2007-2008
• Project: STEVIN demo project

Spelspiek was an interactive and automatic online spelling assistant. It was a chatbot: a robot that is able to simulate a conversation of questions and answers. One could ask Spelspiek something like: “How do you spell bjoetiekees?”, giving a Dutch approximation of the word in question. The chatbot would immediately tell you that the correct spelling is ‘beautycase’. So Spelspiek did not simply correct your spelling errors, it could also provide you with the correct spelling of a word written down as it is pronounced. Words that were not immediately recognized (both neologisms and incorrectly spelled words) were assessed by a spelling expert and added to Spelspiek. Consequently, Spelspiek learned more along the way. Spelspiek was originally a STEVIN demo project (2007-2008). The service was stopped in April 2019.

Succeed

• Acronym for: Support Action Centre of Competence in Digitization
• Time span: 2013-2014

This project promoted the implementation and validation of research findings in the field of mass digitization of text material. Read more (pdf). The INL was co-founder of the IMPACT Centre of Competence in Digitization. It led one of the work packages and was primarily responsible for the selection, adaptation, implementation and evaluation of text processing tools in libraries.

TELRI

• Acronym for: Trans-European Language Resources Infrastructure
• Time span: 1995-1997
• Website: telri.nytud.hu

The objectives of TELRI were similar to those of the PAROLE project. It focused primarily on forming a network between representatives of Eastern European languages and a few Western European ones. For this project, the INL served as the representative for the Dutch-speaking regions.

tranScriptorium

• Time span: 2013-2015
• Website: https://eadh.org/projects/transcriptorium

An enormous number of hand-written historical documents are being placed online by digital libraries across the globe. The unedited digital images only become truly useful once they have been enriched with information. The tranScriptorium project focused on the development of innovative, cost-effective solutions for the indexation, searching, and complete transcription of images of historical hand-written texts. This was done with the help of Handwritten Text Recognition technology. The INL led one of the work packages, and was primarily responsible for the lexica of the various languages and language models used in handwriting recognition. For Dutch, tranScriptorium focused mainly on 15th-century artes manuscripts. Read more.