new
Thesis
Junior
#forschung#wissenschaft#innovation
Master thesis »Extraction of structured materials data through large language models«
Discover 29 Jobs like this
Create an account and we will automatically match you with the most relevant job opportunities.
This text was automatically summarized
Intro
The Fraunhofer IWM is seeking a master's thesis candidate to work on extracting structured materials data using large language models, contributing to innovative research in materials science.
Tasks
- Extract materials information from a literature corpus and image data in a targeted and automated manner.
- Convert extracted information into a structured data format.
- Conceptualize, apply, and refine language models or language model systems that can handle multimodal data.
- Use combinations of different language models with tools for database queries, API calls, or model inference.
- Ensure conformity with a material science application ontology using schema languages like JSON schema.
- Build a bridge between unstructured data sources and existing knowledge graphs.
Requirements
- Studying computer science, computer engineering, computational methods in engineering, or a related subject.
- Familiarity with programming in Python.
- Understanding of the functional principle of language models.
- Initial experience with API endpoints for language model inference.
- Preliminary experience in fine-tuning language models.
- Enjoy working in interdisciplinary teams.
- Systematic and analytical way of working.
- Proficiency in English.
Benefits
- Flexible working hours and mobile working.
- Emergency childcare and parent-child office for family emergencies.
- Employee discounts on events, furniture, clothing, and more.
1 day ago
Mention baito
You like what we are doing? You can support us by mentioning that you found this job on baito.