Building Coffee Commodity Trading Data Warehouse: Decision Making Support Applying Big Data Analysis

Abstract

Coffee is the second-largest soft commodity product, behind oil, traded around the globe and it contributes to the GDP of several countries around the world. The proliferation of the coffee exchange daily and historical market data over the years has created enormous amounts of data. With the potential to increase export value and great domestic demand, the coffee future index plays a significant role in collecting coffee data from major markets worldwide, making reports and forecasts. Current coffee price forecasting is of interest to macroeconomists and transnational trading companies and coffee traders, buyers, and farmers. This research aims to identify and build a data warehouse to help transit the collected data to fuel decision support systems that reveal business intelligence by using my analytical framework to analyze different important parameters in the coffee commodity trading market. Big data is collected from both the London and New York markets from 2010 to 2020 with python code and pre-processed before being analyzed. Collected data was extracted from the active designed database, then transformed to fit my data warehouse structure and loaded into the systems. The Extract – Transform – Load (ETL) process is used to add data to my online analytical processing (OLAP) system. I also identify and visualize parameters that represent different viewing windows and perspectives towards the performance and movement of the coffee trading market for forecasting information to help decision making. The research results will become valuable documents for reference and decision-making support for businesses that trade coffee commodities and for future prediction algorithms.