Expert system for poverty level assessment using the certainly factor method
Keywords:
Artificial Intelligence, Certainty Factor, Expert Systems, Poverty ClassificationAbstract
Poverty is one of the social problems that remains a major concern in national development in Indonesia. The process of determining poverty levels is generally conducted using several socio-economic indicators such as per capita expenditure, education level, sanitation conditions, and access to health services. However, the classification process is often performed manually, causing inconsistencies in determining poor communities.
This study aims to develop an expert system for poverty level classification using the Certainty Factor method. The method is applied to handle uncertainty in the decision-making process by combining several socio-economic indicators into a confidence value for classification results. The research stages include data collection, preprocessing, data transformation, rule-based knowledge construction, inference process, and result analysis. The results show that the Certainty Factor method can be implemented in an expert system to classify poverty levels more objectively and systematically. Based on testing, the combination of low expenditure, medium Human Development Index, and poor sanitation produced a Certainty Factor value of 0.952, indicating a high confidence level in the poor category. The developed system is expected to support decision-making in poverty alleviation programs and social assistance distribution.
