Analytical Dashboard Development for Agricultural Commodities Using Data Mining to Support Food Security

Abstract

Sustainable Development Goals (SDGs) include no poverty, zero hunger, and the achievement of food security. Requirements for food security include 1) adequate availability, 2) stability of availability, and 3) accessibility. A region's food availability can be represented by the potential of its agricultural commodities. The stability of food availability can be seen from time series data on food crop production. Analytical dashboards have become an urgent application platform available to agricultural stakeholders to monitor, predict, map and position commodities as a basis for decision making and establishing policies/programs to support national food security. The aim of this research is to develop an analytical dashboard using data mining. Several data mining techniques were used to build this dashboard. Agricultural commodity predictions use the Autoregressive Integrated Moving Average (ARIMA) and Neural Network (NN) methods. Commodity mapping uses the K – Means and Ordering Points to Identify Clustering Structure (OPTICS) method. The food security model for positioning a region uses Factor Analysis to select significant factors. For food security classification using ordinal Logistic Regression and Random Forest. The best performing methods are implemented into analytical dashboards as needed.