ZHAW Digital Discourse Lab

Unser neues Projekt "Monitoring Vaccination Discourse" startete diesen Februar. In diesem vom Schweizerischer Nationalfonds SNF geförderten Projekt untersuchen wir den öffentlichen Impfdiskurs in der Deutschschweiz von 2000 bis 2025. Philipp Dreesen von der ZHAW Zürcher Hochschule für Angewandte Wissenschaften erklärt im Video, was wir mit dem Projekt erreichen wollen und welchen gesellschaftlichen Nutzen es hat. #SNF #Impfen #Monitoring

🎊🎉 We are happy to announce that our new project FAIR-FI-LD "Moving towards a national FAIR-compliant ecosystem of Federated Infrastructure for Language Data" is funded within swissuniversities' #ORD programme (July 2024-June 2025). FAIR-FI-LD is an Open Research Data project hosted by the University of Zurich with the participation of CLARIN-CH, the Linguistic Research Infrastructure LiRI, ZHAW Digital Discourse Lab and USI Università della Svizzera italiana. The project addresses the crucial need to move towards a federated FAIR-compliant national infrastructure to process, analyse and store various types of language data. Principal investigators: Noah Bubenhofer, Cristina Grisot, Julia Krasselt and Johanna Miecznikowski Read more and stay tunned! https://lnkd.in/e_M42Yn5 #ORD #SSH #FAIR #languagedata

#CfP: PhD-Workshop "Methoden zur Untersuchung beruflicher Sprachpraxis", 02.-04.11.2024 in Berlin

Today, the 1st CLARIN-CH Day will take place in Neuchâtel. The day will be dedicated to Open Research Data, its challenges and opportunities. The morning will be dedicated to short pitches where researchers present challenges and opportunities they discovered in handling research data with respect to Open Science principles. In the afternoon, participants will discuss those challenges and opportunities with experts from different ORD domains (Suzanna Marazza, Brian Kleiner, Simone Mäder, Gerold Schneider, Teodora Vukovic, Christian Futter, Stefanie Strebel). The event seeks to foster exchange and to enable the encounter between researchers and data management experts, as well as legal experts Lab-Member Julia Krasselt is co-organizing the event together with other colleagues from CLARIN-CH. She will present a pitch on challenges regarding the findability of large linguistic corpora such as Swiss-AL in national and European databases.

Im Lab arbeiten wir seit einigen Wochen hochkonzentriert an der Fertigstellung unseres ersten Prototyps der neuen Swiss-AL-Workbench 🚀 . Mit diesem Tool geben wir Wissenschaftler:innen Zugang zu einer wertvollen, mehrsprachigen Textdatensammlung. Die Workbench entwickeln wir im Rahmen der Schweizer Open Science Strategie und durch eine Förderung von swissuniversities und ZHAW Zürcher Hochschule für Angewandte Wissenschaften. Den Prototyp werden wir ab September mit einem Scientific Panel testen. Im Video erklärt Julia Krasselt vom Kernteam des ZHAW Digital Discourse Lab, warum die Bereitstellung von Forschungsdaten der Gesellschaft nützt und welche Ziele das Projekt "Swiss-AL: Linguistische Open-Research-Data-Praktiken für die Angewandten Wissenschaften" verfolgt.

Today, Julia Krasselt from the ZHAW Digital Discourse Lab is presenting research on language use, artificial intelligence and architectural discourse at the conference „AI - Architectural Intelligence“, organized by Andri Gerber and Michael Mieskes from the ZHAW Institute of Constructive Design. 🚀 💡 Julia will show that despite the visual nature of architecture, language analysis is crucial when we want to understand what is meant by important architectural concepts like space and aesthetics. By using techniques also applied in recent AI systems - multidimensional language models, based on neural networks - Julia will give insights into architectural semantic spaces. Crucially, in her talk, she is not using existing language models but a model built for three important Swiss architectural Magazines: Hochparterre, werk, bauen und wohnen Zeitschrift für Architektur and Tec21. In total, the text collection contains 45 Mio. words and covers a time span of almost 45 years. A model based on such a specific, homogeneous dataset allows to pin down expert discourses in architecture.  For a sneak preview, see the video below: It gives an insight into architectural semantic space, showing words used to denote aesthetics.

🏆 Our #SwissAL-CHORD team won second place in the Open Research Data Poster Prize at the Data Science for the Sciences (DS4S) conference held on April 11-12, 2024, in Bern. 🏆   Sooyeon Cho and Dolores Lemmenmeier-Batinić, two members of the SwissAL-CHORD team, presented their poster on the SwissAL-CHORD project at the conference. The conference, organized by the Swiss Data Science Center and Universität Bern, brought together experts in various domains of data science, from statistical data science to machine learning and artificial intelligence, to exchange ideas and explore the latest trends and innovations in open research data.   Congratulations to all the winners and participants, and a huge thank you to the organizers of the DS4S conference! #DS4S #DataScience

We are pleased to share that Sooyeon Cho and Dolores Lemmenmeier-Batinić, two members of the Swiss-AL-CHORD team from the Digital Discourse Lab, participated in the first Data Science for the Sciences (DS4S) conference held in Bern on April 11-12, 2024 (organized by Swiss Data Science Center and Universität Bern) The DS4S conference serves as a platform for experts in various domains of data science—from statistical data science to machine learning and artificial intelligence—to meet and exchange ideas. It was a fantastic opportunity to explore the latest trends and innovations in open research data as well.  Our team had the privilege of presenting the Swiss-AL-CHORD project, engaging in stimulating discussions that sparked new ideas and potential collaborations. We're excited about the connections we've made and the impact these will have on our work moving forward. A big thank you to the DS4S organizers for a wonderfully productive conference!

𝐃𝐨 𝐲𝐨𝐮 𝐨𝐟𝐭𝐞𝐧 𝐟𝐢𝐧𝐝 𝐲𝐨𝐮𝐫𝐬𝐞𝐥𝐟 𝐰𝐨𝐧𝐝𝐞𝐫𝐢𝐧𝐠 𝐰𝐡𝐞𝐫𝐞 𝐩𝐞𝐨𝐩𝐥𝐞 𝐠𝐞𝐭 𝐭𝐡𝐞𝐢𝐫 𝐢𝐧𝐟𝐨𝐫𝐦𝐚𝐭𝐢𝐨𝐧 𝐫𝐞𝐠𝐚𝐫𝐝𝐢𝐧𝐠 𝐯𝐚𝐜𝐜𝐢𝐧𝐚𝐭𝐢𝐨𝐧𝐬? 𝐓𝐡𝐞𝐧, 𝐲𝐨𝐮 𝐦𝐚𝐲 𝐛𝐞 𝐢𝐧𝐭𝐞𝐫𝐞𝐬𝐭𝐞𝐝 𝐢𝐧 𝐭𝐡𝐞 𝐃𝐢𝐠𝐢𝐭𝐚𝐥 𝐃𝐢𝐬𝐜𝐨𝐮𝐫𝐬𝐞 𝐋𝐚𝐛'𝐬 𝐧𝐞𝐰 𝐩𝐫𝐨𝐣𝐞𝐜𝐭, "𝐌𝐨𝐧𝐢𝐭𝐨𝐫𝐢𝐧𝐠 𝐕𝐚𝐜𝐜𝐢𝐧𝐚𝐭𝐢𝐨𝐧 𝐃𝐢𝐬𝐜𝐨𝐮𝐫𝐬𝐞." 𝐓𝐡𝐢𝐬 𝐩𝐫𝐨𝐣𝐞𝐜𝐭 𝐚𝐢𝐦𝐬 𝐭𝐨 𝐚𝐧𝐚𝐥𝐲𝐳𝐞 𝐒𝐰𝐢𝐭𝐳𝐞𝐫𝐥𝐚𝐧𝐝'𝐬 𝐯𝐚𝐜𝐜𝐢𝐧𝐚𝐭𝐢𝐨𝐧 𝐝𝐢𝐬𝐜𝐨𝐮𝐫𝐬𝐞, 𝐚𝐧𝐝 𝐢𝐭𝐬 𝐟𝐢𝐫𝐬𝐭 𝐩𝐡𝐚𝐬𝐞 𝐢𝐬 𝐜𝐮𝐫𝐫𝐞𝐧𝐭𝐥𝐲 𝐢𝐧 𝐩𝐫𝐨𝐠𝐫𝐞𝐬𝐬!   Imagine: You're at a big Easter family gathering, and everyone starts discussing whether to get the flu shot. Some family members are in favour of it, sharing stories of how skipping the shot made them really sick. Others are unsure, citing possible side effects they read about online. This type of conversation, where everyone shares their opinions and experiences, is similar to what happens on a larger scale in Society in the area of public communication.    Public communication about vaccines can promote vaccine acceptance or increase hesitancy and refusal. Therefore, it is crucial to understand how the media reports on vaccines, how the topic is discussed in politics, and how people talk about vaccinations in public. What is the current discourse regarding vaccines and vaccinations in Switzerland?    On February 27th, we started a new project called "Monitoring Vaccination Discourse" to analyze precisely that. The project's first phase, modelling the text corpus, entails identifying the key actors in the vaccination discourse in the German-speaking part of Switzerland from 2000 to 2025. Along with media, politicians, and Public Health actors, information portals and internet forums aimed at laypeople are also important sources of information and, thus, carriers of social knowledge about vaccinations.    To identify these sources, we   analyzed which websites people can typically find online when searching for information about vaccinations. We used different search queries to cover general and specific questions such as health information, vaccination, vaccination+allergy, or vaccination+immunosuppression.   We also looked for websites that provide information about diseases against which people in Switzerland can be vaccinated, such as measles or TBE. This resulted in an extensive list of actors whose relevance we will further investigate in terms of additional aspects. However, it's important to note that this type of modelling may limit the diversity of public discourse. Therefore, we plan to involve practitioners in the project to ensure a more comprehensive approach.    How do you usually search for information about vaccinations? Which sources do you trust? Tell us in the comments! 

Last but not least, we are examining the fourth principle of FAIR Data, which is #reusability:  The main aim of FAIR is to optimize data reuse by providing well-described metadata and data that can be replicated and combined in different settings.   How can we ensure that data on Swiss-AL is reused?   Most importantly, we make Swiss-AL reusable for researchers in applied sciences that are interested in language data but lack explicit (corpus) linguistic training. Existing tools for language data analysis tools limit access to a small group of linguistic experts, excluding our target group. Additionally, legal restrictions prevent us from storing Swiss-AL on a conventional data repository.   To that end, we are implementing a dedicated browser-based analysis platform on Swiss-AL we call the "workbench".   This workbench will offer easy access to Swiss-AL and include analysis functions that are of interest for researchers in various fields, such as the social sciences, health sciences, law, and architecture.   In order to improve reusability, the platform will allow researchers to access their data in a way that enables a wide range of linguistic analyses to be performed directly on the platform. But also download of data will be possible, e.g. to continue an analysis in a different tool.