Developing Forecasting Model of Vegetable Price based on Climate Big Data

Big data is one of the most discussed topics in recent economic and business sectors with explosive applications of information and communication technologies (ICT). The object of this study is to develop a forecasting model based on a big data processing. This study focuses on the forecasting of vegetable price considering climate factors as one of major big data associated with the agricultural field. Onion and napa cabbage in Korea are selected as target products. Price forecasting models are constructed by a Bayesian structural time series (BSTS) and a vector autoregression (VAR) models. Both models introduce climate factors of temperature, precipitation, sunshine duration, and the lowest temperature in chief producing district for onion and napa cabbage. Results show that, for onion price, BSTS is more appropriate for the short-term price forecast, and VAR for the long-term. For napa cabbage prices, both BSTS and VAR show similar patterns in price forecasting. However, BSTS predicts price relatively lower than VAR does. We conclude that it is necessary to consider big data concerning climate factor in forecasting vegetable price and to develop various models across agricultural products with their growing environment.