GIS-based and Naïve Bayes for nitrogen soil mapping in Lendah, Indonesia

Abstract

Rice or Oryza sativa L. is the staple food of Indonesian society. Currently, the demand for rice in Indonesia is increasing, while the level of rice production is decreasing. Therefore, alternative technologies are needed to assist the community in increasing rice production. One of the important factors in the growth of rice is the quality of the soil that contains high levels of nitrogen for photosynthesis, protein formation, acid formation, and accelerating plant growth. This research was applied for soil nitrogen mapping with a smart prototype using the TCS3200 sensor combined with Naive Bayes algorithm and GIS (Geographical information systems). This system is carried through the Wemos D1 Mini microcontroller of TCS3200 sensor for reading data and sending the data to the web ceerduad.com. Geographical information systems (GIS) was developed to mapping area and to obtain training data for the Naïve Bayes Algorithm which the researcher took 20 soil samples in Lendah sub-district. From the experiment results, a prototype of soil nitrogen content using the TCS3200 sensor can measure soil nitrogen levels with an accuracy of 87.5% and sending the sensor data to the web server. Implementation of GIS in this research was successfully mapping low, medium, and high levels of nitrogen.