The Use of Toulmin's Argumentation Model in Solving The Drug Conflict Problems

The field of argumentation in Artificial Intelligence (AI) has witnessed a great increase an important cognitive to deal with uncertain information and conflicting opinions. This has led to a number of interesting lines of research in this field and related fields, giving rise to computational models of the argument as a promising research field. The remedies conflict problem is considered one of the challenges in the field of medicine the world. This paper makes use of Toulmin's argumentation model to deal with conflicting problems within the medicine field. In addition, inference rules were used for associating a patient's symptoms and patient history(premises) with remedies use, eventually leading to medications diagnosis for patient (claims). After that, several remedy features are used to compete for the support and the attack (pros and cons) for each remedy item. A decision is made during the qualifier phase in Toulmin's model about whether or not the drug should be used based on the highest value of support or attack. The dataset consists of 200 patients as samples for two heart diseases (hypertension, angina pectoris). It is collected from the Iraqi educational hospitals, annotated by a team of experts working in the medical field. The performance achieved in the proposed model in hypertension and angina pectoris diseases were 78% and 83%, respectively, using the confusion matrix method.