Preparation of Smart Card Data for Food Purchase Analysis of Students through Process Mining

Abstract

This research emphasizes on preparation of data collected from students of a primary school who have to use digital food cards (from a variety of food vendors inside or in the vicinity of their school's campus) in order to purchase any food product(s) within (or even out of) the allowed study periods. To do this, they datasets initially were converted into a CSV format file in such a way to be supported in the Disco Fluxicon environment, which is a process mining tool and platform. Accordingly, the research includes the following steps: 1) Data collection and data gathering, 2) Data cleansing and data filtering, 3) Data conversion and exporting the data in appropriate format. The proposed method applied in this experiment was based on real event logs from an authentic primary school in Thailand.