<![CDATA[The current irrigation system is still manual, resulting in suboptimal water use and water distribution that does not meet land requirements, causing a lot of water to be wasted. Irrigation control is also ineffective because Internet of Things technology is only capable of monitoring without automatic decision-making. This situation causes social problems related to water scarcity and slows down the progress of sustainable agriculture. The purpose of this study is to create an Internet of Things-based Sistem Irigasi cerdas system that can monitor agricultural environmental conditions in real time and use the C4.5 algorithm to classify irrigation needs. The research was conducted in the rice fields of Jabong Village with direct observation of agricultural land. Data was collected through DHT22, capacitive soil moisture, and raindrop water sensors with a total of 100 records. The analysis stages included determining the root node of all attributes, calculating entropy, information gain, and gain ratio. The attribute with the highest gain ratio was used as the root node, then the data was split to form leaf nodes. After the tree was formed, pruning was performed to avoid overfitting. The test results using RapidMiner tools showed an accuracy of 90.00%, which is classified as “good,” so this system can be a breakthrough from conventional agriculture to modern agriculture. This study is still limited to one user, three parameters, and the C4.5 algorithm.]]>
