Master project: NLP for analyzing open-ended questions in hospital questionnaires
Suzan Verberne
See below the description of a thesis opportunity for a MSc Thesis project at the LUMC. Feel free to contact Marieke directly through the email address below, or contact me if you have questions on supervision.
Best regards,
Natural Language Processing for analyzing open-ended questions in hospital questionnaires
Contact persons: Suzan Verberne and Marieke van Buchem (PhD candidate in the LUMC), M.M.van...@lumc.nl
Patients’ experiences are important to improve healthcare and are often collected using a combination of closed- and open-ended questions. Open-ended questions give patients the opportunity to share their unique perspective and thus provide more information than closed-ended questions. However, the analysis of open-ended questions is hard to automate and thus takes up a large amount of time. Therefore, the LUMC is working on using Natural Language Processing (NLP) to automate the analysis of open-ended questions for an efficient and patient-centered approach to improve healthcare.
In the past year, we have created a first NLP-pipeline that extracts the most important subjects from patients’ experiences using topic modeling. The topic model shows promising results, but we believe there is room for improvement. During this internship you will work on creating a new and improved NLP-pipeline, using language models, topic modeling, and aspect-based sentiment analysis. You will work closely with the end users and focus on delivering a pipeline that will be implemented in practice. The data you will be working with consists of real patient experience data, collected in the LUMC and several other hospitals over the past years.
--
Leiden Institute of Advanced Computer Science
http://liacs.leidenuniv.nl/~verbernes