In this series we showcase high-quality and innovative research that uses CLARIN tools and resources. These impact stories illustrate the huge variety of disciplines that use the CLARIN infrastructure, highlight the excellent research linked to it, and demonstrate the wider impact that CLARIN and the social sciences and humanities have on broader societal issues.
Explainable AI - Political Orientations in Slovenian Parliament
Read how machine learning models were developed and explained in order to understand the language used by Slovenian members of parliament associated with different political leanings.
MATEO: Easy and Accessible Machine Translation Evaluation
MATEO is a new, user-friendly tool for both experts and. non-experts that answers a growing need for easy and accessible evaluation of machine translation.
State-of-the-Art Speech Recognition for Oral Histories
A team from CLARIN’s Czech node has developed a system that uses speech recognition and NLP technologies specifically for oral history archives.
Gender in Poland’s Presidential Election Campaigns
Using CLARIN tools, this project reconstructs and analyses how the notions of sex and gender featured in the 2015 and 2020 Polish presidential election campaigns.
Chatbots and Copyright: CLARIN Café Addresses Key Aspect for DH
In this latest CLARIN Café, CLARIN’s legal experts explore the legal implications of working with or using AI-generated texts, and the related copyright issues.
Networks of Power - Gender Analysis in European Parliaments
Using the ParlaMint dataset, this project examines different aspects of power in three European parliaments, with a particular focus on gender distribution in the debates.
ParlaMint - A Resource for Democracy
This project explores the public discourse on migration and migrants in Italy and the UK, and shows how this may impact public opinion on the topic.
Voices from Ravensbrück: Multilingual Oral History
This project brings together oral interviews by survivors of Ravensbrück concentration camp and presents a unique opportunity to compare these historical sources.
Ukrainian History Course in Response to the War in Ukraine
Read more about the distant learning course 'Ukrainian History', which was developed in response to the invasion of Ukraine and was supported by CLARIN.
Re-evaluating Child Language Assessment Measures
Dr Nan Bernstein Ratner presents her latest project and discusses how spoken language data is increasingly being used by the ASR industry.
Discovering Slovenian Language Structure Using Corpora
Jakob Lenardič’s PhD project combined theoretical and corpus linguistics to explore the subtle charateristics of Slovenian language structure.
Open Language Resources for Smarter Artificial Intelligence
Watch Kaja Dobrovoljc's recent presentation at ESFRI's 20th Anniversary Conference, where she underlines the importance of open language resources for AI.
Navigating the GDPR with Innovative Educational Materials
‘Privacy in Research: Asking the Right Questions’ is an engaging response to the challenges faced by researchers working with sensitive data.
Donate Speech Database to Boost Development of AI Applications
This project has collected around 4000 hours of colloquial Finnish speech to accelerate the development of language-based AI applications.
IceTaboo: Offensive Word Database with Commercial Application
The IceTaboo database is already being used as part of an automatic proofreading software by an Icelandic online news website.
Xenophobia on Greek Twitter during and after the Financial Crisis
Following up on an earlier study, this project investigates whether economic and political changes have affected the public attitudes expressed on Twitter.
Tracing Language Change with a Monitor Newspaper Corpus
This study traces the linguistic changes that occurred in the Norwegian language during the first wave of the COVID-19 pandemic.
Stories in Motion: Oral History as Sustainable Data in Urban Settings
This project is developing a model for archiving and (re)using oral histories, extending the collections' impact for research and the community.
Helsinki Digital Humanities Hackathon DHH21
The hackathon uses the ParlaMint 2.1 dataset to compare how four different national parliaments responded to the COVID-19 pandemic in their debates.