Dynamic Inquiry System for Hospital Department Recommendation

Abstract

This research project explores the application of artificial intelligence (AI) in developing a semi-automatic department recommendation system via a web-based service, aimed at improving the ease of hospital appointment scheduling for patients. The system comprises two main components: an inquiry module and a department prediction module. Basic health information and patient-reported symptoms are collected and analyzed using AI models to predict the most appropriate hospital department and dynamically generate follow-up questions. The input dataset consists of symptom records labeled with diagnosed diseases, each mapped to its corresponding department. We investigated and developed multiple models to support both the inquiry and prediction tasks, including Extreme Gradient Boosting (XGBoost), a long short-term memory (LSTM) network, which is a recurrent neural network (RNN), and embedding-based deep learning models. Model performance was evaluated using both partial input (1–5 symptoms) and full input (all symptoms). XGBoost demonstrated the best overall performance, achieving an accuracy of 99% with the full feature set. This research project ultimately aims to improve hospital service efficiency by guiding patients to the appropriate department, reducing wait times and confusion. It also equips medical staff with preliminary patient information, enabling better preparation and more accurate treatment planning.

Ähnliche Arbeiten