Vacant job

University of Gothenburg meets society's challenges with diverse knowledge. With 58,000 students and 6,800 employees, the university is a large and inspiring workplace. Strong research and attractive education attract researchers and students from all over the world. With new knowledge and new perspectives, the University of Gothenburg contributes to a better future.

The Department of Swedish, Multilingualism and Language Technology at the University of Gothenburg conducts research and education in various subjects at undergraduate and advanced levels. At the department, research and doctoral education are conducted in the field of language technology within the department's research unit, Språkbanken Text. Language technology researchers actively participate in interdisciplinary projects, collaborate with the surrounding society, other departments, and various actors in society. The research activities within Språkbanken Text revolve around creating, analyzing, refining, and providing linguistic resources that can be used in research and teaching. We are now seeking one to two assistants, placed at the Department of Swedish, Multilingualism and Language Technology during the summer/autumn of 2026.

Job Responsibilities

The main focus of the employment covers three types of tasks:

• Segmenting ("cutting") long audio recordings into segments adapted to the boundaries of each test.
• Reviewing and correcting automatically generated transcriptions. This may involve correcting misspellings, completing missing word and sentence parts through orthographic transcription, and correcting incorrect word segmentations.
• Using manual reference assessments to compare and correct the automatically generated scoring of cognitive tests. The scoring has been designed and implemented using embedding-based language models, with Swedish as the target language.

The material comes from recordings within the H70 study, a long-term Swedish population study on aging and health. The study includes recurring examinations of older individuals and includes, among other things, digitally recorded neuropsychological tests.

All work tasks will be introduced briefly. Therefore, previous experience in working with, coding, and correcting similar material is highly meritorious. It is also important that you are well acquainted with the design, structure, and assessment principles of the tests in question. Examples of tests you will work with include phonological word fluency tests, such as F-A-S, as well as the Boston Naming Test (BNT), and similar tests. These tests are used to investigate various cognitive functions, including language, vocabulary, and memory. They have long been used in both research and clinical settings to study linguistic and other cognitive difficulties with various underlying causes.

For audio processing and annotation, Nextcloud should be used as a secure platform for storing, synchronizing, and sharing files. The actual processing, however, should be performed locally in PRAAT, a computer program used to analyze, process, and annotate audio files. Since PRAAT is normally not run as a Nextcloud app, the solution is to store and synchronize the material via Nextcloud, but to open and process the files locally in PRAAT.

Please note that a confidentiality agreement will be established and used during the recruitment process, in accordance with the University of Gothenburg's regulations.

For further information, qualifications, and eligibility requirements, as well as to apply for the position, see the full advertisement at https://www.gu.se/jobba-hos-oss/lediga-anstallningar (https://www.gu.se/jobba-hos-oss/lediga-anstallningar).