A supervised learning approach for recommending medical specialists in the healthcare sector for the Afaan Oromo context

Abstract In healthcare institutions, an automated system plays a critical role by enhancing patients’ satisfaction with service delivery. This paper focused on the development of a model that assists patients in finding the appropriate medical specialists in Afaan Oromo. To do this, text preprocessing tasks were applied to the data to remove unnecessary texts, punctuation, and numbers, as they would be suitable for the training model. A feature extraction task is applied to obtain a standard Afaan Oromo health dataset using TF-IDF. We used supervised learning algorithms such as logistic regression, random forest, multi-layer perceptron, decision trees, Bi-LSTM, and K-NN algorithms for experimental purposes. Evaluation measures were used in comparing the performance of the model with seven specialist classes on the labeled dataset. In the comparative analysis, the result reveals that Bi-LSTM performed well, achieving an equal value of accuracy and F1 score, which is 0.9708. Based on the experimental results, a user interface was developed for the proposed method, and the highest-outperformed model is to detect the symptoms and predict the appropriate specialists.