Artificial Intelligence
[Closed]
Work description
To develop the named entity model, some baselines of existing models can be tested initially. Then, if the results of the pre-trained model are satisfactory, they will be used to support entity labeling. Finally, a model using a Transformers model will be trained to identify named entities. For the second objective, the student will identify whether there are existing open-source models for the task. If such models exist, the first task will be to apply them as baselines. Next, we will proceed to label a small sample of data, which will include both images and text. Finally, based on this sample, we will train a multimodal model on the data. If the data is insufficient, more data will be labeled. For the third objective, experiments with generative models will be required to generate suggestions based on the output of the model from the second objective. The fellow will be able to identify specific models for medical language, if necessary.
Academic Qualifications
Bachelor's degree in Computer Science, Artificial Intelligence, Data Science or equivalent.
Minimum profile required
Quality academic training in Computer Science, Artificial Intelligence, Data Science, or equivalent. Knowledge of LLMs/NLP and multimodal models.
Preference factors
NLP thesis topic or experience with NLP, some knowledge or experience with the biomedical domain.
Application Period
Since 25 Sep 2025 to 08 Oct 2025
[Closed]
Centre
Artificial Intelligence and Decision Support