<![CDATA[Environmental health is one of the important components that are very influential for human life and health. A clean and healthy environment will make every individual around him comfortable and improve the quality of his life. Toilets are part of an environment that is very important. The poor quality of the toilet facilitates the transmission of bacteria and the development of germs. Based on these problems a system was created that could be used to classify the quality of toilets based on several parameters. In this study the parameters used in comparing the quality of the toilet are odor and temperature on the toilet. Where for odor parameters consists of ammonia gas, carbon dioxide, and carbon monoxide. These gases are gases that are often produced in human activities when on the toilet. The process of determining toilet quality through ammonia, carbon monoxide, carbon dioxide, and temperature is obtained from reading two MQ135 sensors and DHT11 sensors by the Arduino Uno microcontroller using the Naive Bayes method. The use of the Naive Bayes method was chosen as a decision making technique for toilet conditions because it has very good accuracy where the class of toilet type classification has been known from the beginning. From the results of several tests the reading of two MQ135 sensors has a very high correlation with the output voltage. Where for ammonia reading has a correlation of 99.13%, carbon monoxide has a correlation of 99.66%, and carbon dioxide 99.22%. Whereas for the temperature reading of the DHT11 sensor it has a presentation error of 0.502%. Furthermore, in testing the Naive Bayes method system with a total of 55 training data and test data as many as 25 data, obtained an accuracy of 96% with an average computing time of 4.59 seconds.]]>
