Model Creation in QPR ProcessAnalyzer

Model Creation

In order to create functioning data models for Process Mining the minimum requirement is that you have an event log which contains the following information:

• Case ID - An unique identifier which connects the series of events together, eg. Purchase Order number, invoice number etc.
• Event name - The name of different process steps, eg. Delivery: Goods Issue, Invoice Receipt, SO Item Created etc.
• Timestamp - A registered time of the occurrence of an event
• Attributes - Any additional information attached to a Case ID, eg. Company Code, Customer information, Order status information etc. Attributes are not mandatory for creating functioning models but the more attributes there are in the model the higher quality analysis can be done.

Based on this information QPR ProcessAnalyzer creates a flowchart which with combination of other available tools lays the foundation for the start of Process Mining. Most common ways of creating models in QPR ProcessAnalyzer are using the webUI to import data and more advanced techniques such as SQL to import data directly from various data sources.

Data Import via QPR ProcessAnalyzer UI ​

Typical scenario for using​

An organization has some ETL platform (potentially with access to multiple IT system / data lakes) in use which provides CSV file as output​

• Employee uploads CSV files via web browser to QPR ProcessAnalyzer​
• New process mining model is created in couple of minutes!​
• Limitations:​
  • CSV data must be in the Process Mining data format as no data editor available​
  • CSV size max. 230mb

QPR ProcessAnalyzer MS Excel add-on

Typical scenario for using​

QPR ProcessAnalyzer user has MS Excel add-on installed and wishes to do some data manipulation with MS Excel standard functionality​

Guided wizard to help user on the steps to create new model​

New process mining model is created in a couple of minutes​

No SQL knowhow needed and easy data transformation/manipulation​

Limitations:​

CSV data must be in the Process Mining data format​

QPR Integration platform

Typical scenario for using​

An organization would like to dynamically connect with IT systems to schedule frequent data queries​

Developer uses QPR’s readily available standard connectors to query data from IT system & create process mining model​

The standard connectors may to customized fully for the specific needs e.g. new case attributes or process steps ​

​

Limitations:​

No limits regarding data format or size​

Requires SQL knowledge

Direct ODBC connection to database​

Typical scenario for using​

An organization has S/4 Hana (or similar) advanced data repository with very high computing power ​

Developer uses QPR’s readily available standard connectors to query data from IT system​

The standard connectors may to customized fully for the specific needs e.g. new case attributes or process steps ​

Always up-to-date process mining model and no separate scheduling required because data is retrieved from the source system once the model is opened from QPR ProcessAnalyzer​

​

Limitations:​

All the data must be queried with single ODBC query. This means that a single query might be extremely long and complex to be managed.​

Requires SQL knowledge​ Max 1 million rows (Excel constraint)