Navigation auf uzh.ch
Navigation auf uzh.ch
DSI Infrastructures & Labs are shareable infrastructures or structural vessels for creating collaborative research environments related to digital transformation. Prof. Dr. Viktor von Wyl briefly introduces Talk2UZH («talk to us»), one of the projects in this series.
What is the Talk2UZH Lab?
Our experience from studies with chronically ill people shows: Study participants would rather tell their story than fill out monotonous questionnaires. New technologies such as large language models can simplify data collection as well as efficient data processing and analysis – even in large study populations. The Talk2UZH Lab consists of an interfaculty team with an interest in qualitative interviewing and language analysis. Our Lab aims to make speech recognition accessible to UZH researchers through selected tools in an easy and safe framework for research and teaching. Our initiative is closely linked to the Population Research Center and the Linguistic Research Infrastructure (LiRI).
What are the stumbling blocks in using speech recognition in research?
The biggest stumbling blocks are ease of access and usability, data privacy and control, and lack of «best practices» in collecting and analyzing voice and free text data. The Talk2UZH Lab is striving to improve in these three areas. Our speech recognition software is integrated into intuitive application software, runs locally on servers at the University of Zurich, and does not store sensitive data. In addition, Talk2UZH Lab team members have extensive experience in collecting, processing, and analyzing speech data, which we aim to share and refine with other researchers.
What does the pilot software pay particular attention to?
The speech recognition software should be easy and safe to use, protect privacy, and function in multiple languages. Equally important is a solid understanding of the opportunities and risks of using speech recognition in research and teaching. Therefore, the Talk2UZH team is planning an application project, which will be conducted together with members of the Population Research Center, LiRi and others who are interested. The experiences from this project will be systematically reviewed and published afterwards from different perspectives – such as usefulness, ethics, efficiency.
Learn more about Talk2UZH here.
All projects of the series «DSI Infrastructures & Labs» can be found here.
Prof. Dr. Viktor von Wyl heads the Digital & Mobile Health research group at the Institute for Implementation Science in Health Care. He is also scientific director of the Swiss Multiple Sclerosis Registry at the Institute for Epidemiology, Biostatistics and Prevention and a DSI Professor. Educated in epidemiology and health services research, he has participated in the design and analysis of large health observational studies involving the general population (Corona Immunitas) or individuals with multiple sclerosis (Swiss MS Registry). Building on this experience, the Digital & Mobile Health group aims to develop next generation participatory health observational studies. These new study concepts should on the one hand integrate existing data (e.g. administrative data) as efficiently as possible and combine it with newly collected clinical and survey data, as well as with data from wearable sensors (e.g. activity trackers) and speech recognition and analysis.
