System Models for Forecasting Business Models: A Case Study of Agricultural Online Wholesale Marketplace

Abstract

The objective is to develop the system model in order to forecast the online supply business model for agricultural product by employing decision tree method. There are 3 divided parts in the research methodology. The first part is the data collection, in which the data collected for this study and analysis is secondary data from reliable sources. The second part is the stimulation of sets of data on the visual trading environment in accordance with the information from the first part. The third part is to create the model for forecasting. The data is provided for WEKA in order to create the model by using Tree J48 (C4.5) algorithm. The classifying process will be conducted by Tree J48 (C4.5) and the effectiveness evaluation of the forecasted result will be calculated by Confusion Matrix. According to the study, its objectives are accomplished by 1) the model has been able to forecast the sale condition in order to provide agriculturalists the most profits and the result of Accuracy, Precision, Recall, and F-Measure are 77.33%, 0.79, 0.77, and 0.78, respectively. 2) the model has been able to forecast the buy condition in order to provide an opportunity for middlemen to receive the lowest product prices and the results of Accuracy, Precision, Recall, and F-measure are 80.44%, 0.80, 0.80, and 0.80, respectively.