Proposed a Solution Based on Functional Dependency Graphs for Transactional Fact Table Error Correction

Process mining refers to the set of techniques used to mine the event logs of process-oriented information systems that manage and execute processes to gain hidden knowledge about an organization's actual business processes. Data warehouses are a form of database that stores data from different sources. Data warehouses contain three types of tables: transactional fact table, periodic fact table, and cumulative fact table. The transactional fact table, contains a frequent error in the mapping between the primary key, cumulative fact table measure and dimension level granularity. This paper proposed a solution based on functional dependency graphs to correct the frequent errors in the transactional fact table.