Sia et al. (2024) developed an AI-based disease prediction chatbot system with a promising accuracy result of 90%. They used four machine learning algorithms to build a prediction model that can recognize potential diseases based on reported symptoms. Here is the performance of each algorithm:
Table 1. Performance of machine learning algorithms in the disease prediction chatbot system
| Algorithm | Accuracy |
| support vector machine (SVM) | 92.24% |
| random forest | 92.23% |
| k-nearest neighbors (KNN) | 91.57% |
| artificial neural network (ANN) | 91.52% |
From Table 1, the SVM algorithm demonstrates the highest accuracy results. These findings showcase the model’s capability to effectively identify patterns and correlations between diseases and symptoms.
The developed chatbot is integrated with long short-term memory (LSTM) and natural language toolkit (NLTK) for natural language processing and implemented through the Telegram platform to create an accessible and user-friendly interface. Although the results of this research are highly encouraging, Sia et al. (2024) emphasize the importance of addressing data privacy issues and the security of users’ health information. They also recommend further development, including expanding the dataset by adding information sources such as medical history, lifestyle factors, and genetic data.
This innovation represents a significant step forward in the application of AI in the healthcare sector, with great potential as an early diagnostic aid for medical professionals and the general public.
Reference:
M. Sia, K.-W. Ng, S.-C. Haw, and J. Jayaram, “Chronic disease prediction chatbot using deep learning and machine learning algorithms,” Bulletin of Electrical Engineering and Informatics, vol. 14, no. 1, pp. 742–751, Nov. 2024, doi: 10.11591/eei.v14i1.8462. Available: https://doi.org/10.11591/eei.v14i1.8462