Transactions Graph is the network model created to generate list of predictions about future spending. It shows the unprocessed model of dynamic relationships between transaction items and their effects changing overtime. Its topology intensifies or de-inensifies depending on the quality and force of the relations. Each node represents a transaction and connected to each other through categories and dates. Thickness of connecting lines shows the total spending amount of the paired transactions. At a given timeframe, the last day's purchases and the two weeks of past spending are shown.