Call for societal or commercial project proposals 2022

CLARIAH wants to maximize its potential as a centre for innovation. The CLARIAH infrastructure offers a range of uses not only for academic institutions, but also for commercial companies and social partners.

Wider relevance

It is of great importance that the CLARIAH infrastructure is not only used by scientific researchers, but also plays a role outside scientific research and can have societal and/or commercial impact. This contributes to the long-term sustainability of the CLARIAH infrastructure. It is also one of the Key Performance Indicators formulated by ESFRI for ESFRI Landmark infrastructures such as CLARIN and DARIAH, and consequently CLARIAH in the Netherlands.

Research proposals

Dutch researchers can apply for funding of a small research project that helps to achieve societal and/or commercial impact. Conditions for such projects or activities are:

• It concerns small exploratory projects that may carry out preliminary research in order to arrive at a larger externally financed project.
• They address societal and/or commercial themes or problems and promote the impact of CLARIAH in these areas.
• They align with themes and research questions from the humanities, in particular with core themes and focus areas of the CLARIAH work packages.
• The theme is in line with themes from the Top Sectors, the Dutch National Science Agenda, the Open Science movement (i.e. the EOSC-Hub), or the Sectorplan Social Sciences & Humanitiesof the Ministry of Education, Culture and Science.
• There is co-financing from a social partner, for at least half of the budget, in-kind or in-cash.
• The maximum contribution per application from CLARIAH is € 10K.
• A researcher or research group within CLARIAH makes an active contribution to the project. (CLARIAH can help to connect proposals to existing research).

The total budget available for these projects is €70K.

Submitting a proposal

• Proposals can be submitted by email (in Dutch or English) in PDF format. At the least fill in all parts of the proposal form.
• Send your proposal to clariah@huygens.knaw.nl. You can also contact the CLARIAH board via this e-mail address to discuss the possibilities for collaboration with one of the existing projects.
• The deadline for submission is March 1, 2022.

Examples of previous projects

Say no more! Natural Language Processing for diagnosis and screening in psychiatry

• Duration: November 2020 - November 2021
• PI: Drs. Janna de Boer, MD, UMC Utrecht
• Funding: Sunovion, CLARIAH

Much research is being done on the value of spoken language analysis in psychiatry, as speech is an important source of information about how someone is doing. In addition, many psychiatric disorders are characterized by (mild) language and/or speech disorders. Until now, such analyzes are very labor intensive, as all conversations have to be transcribed manually.

With this project we want to use CLARIAH's automatic speech recognizer (ASH) to analyze whether speech from patients with a psychiatric disorder can also be accurately recognized automatically. At present it is not yet clear whether the speech of people with a language and/or speech disorder can be recognized accurately enough to be used for clinical purposes such as diagnostics or prognosis.

ASR4PESP: Automatic Speech Recognition for Primary Education Speaking Proficiency

• Duration: November 2020 - November 2021
• PI: Dr. Catia Cucchiarini, Radboud University Nijmegen
• Funding: CITO, CLARIAH

Assessing speaking skills in primary education is not easy because teachers have to pay attention to many aspects (content and grammar) in one conversation. As a possible solution, CitoLab has developed Beeldverhaal (picture story), an app that invites students to tell a story based on four pictures. With Automatic Speech Recognition (ASH) the spoken story is converted to text so that the teacher can later review it in multiple aspects to better assess speaking skills.

Because ASH of Dutch children's speech is still problematic, this project improves the quality of ASH of Dutch children's speech in order to make it suitable for educational applications such as Beeldverhaal.

SASTA (Semi-Automatic Analysis of Spontaneous Language)

• Duration: September 2019- September 2020
• PI: Prof. Jan Odijk, University of Utrecht
• Funding: Association for Clinical Linguistics, Language Technology Foundation, CLARIAH

Clinical linguists like to use the analysis of spontaneous language sessions with patients to diagnose language deficits or abnormalities (eg, aphasia). The analysis of the data from such sessions has so far been done entirely by hand. This is very labour-intensive and is therefore often not done or only partially.

The University of Utrecht, together with the Vereniging voor Klinische Linguïstiek (VKL) is investigating whether this process can be partially automated to make this process more efficient. To this end, a research prototype of a new application is being developed (SASTA) that is suitable for the semi-automatic assessment of language development according to different assessment methods.

This new application:

1. reuses modules from GRETEL4 (i.e. the Alpino automatic parser and GrETEL's query system) ;

2. adds new modules for aspects not or not fully covered by Alpino (e.g. for compounds, certain forms of spoken language and all kinds of deviant language use).

SASTA uses specifications formulated as a set of queries that implement language assessment methods such as STAP, TARSP, and ASTA.