Process mining and discovery is a rapidly growing market segment that empowers organizations to gain visibility to how their processes work, identify root causes of process inefficiencies, and make informed business decisions to optimize process execution.
A key requirement for effective process mining is the ability to access event logs from various systems of record, as they actually occur. Such event log data "is often considered as the 'new oil' and data science aims to transform this into new forms of 'energy' as insights, diagnostics, predictions, and automated decisions".1 Data transformation is a necessary process to transform the raw data (the "new oil") into meaningful insight ("energy").
What is ETL?
Data transformation commonly referred to as Extract, Transform, and Load (ETL) is a process to transform data from multiple systems and applications for analysis purposes. This process includes normalization, cleansing, deduplication, and formatting as a part of a multi-stage data transformation process, before moving the data into a data warehouse for further analysis.
The application of ETL is particularly useful to help businesses analyze structured data relating to business performance using online analytical processing (OLAP) tools. To ensure meaningful analysis, the transformation of source data is essential and traditionally requires IT skills with SQL/NoSQL, scripting, and data mapping expertise.
ETL vs ELT
When it comes to common business users exploring new technologies like process mining or other, traditional ETL tools are not as useful. A variant approach is what is referred to as ELT-Extract, Load, and Transform, whereby the extracted data is immediately loaded into an analytical application such as a process mining platform where data transformation occurs, thereby delivering faster time to value without the need for time-consuming and expensive data transformations.
The value of ELT lies in its ability to support transforming very large volumes of data, including real-time streaming, leveraging the power and scalability of the cloud infrastructure to be consumed by business users.
Andrew Pery, Digital Intelligence Consultant, ABBYY
ELT is particularly useful to support process mining of event data logs associated with case-based processes with a high degree of variability, such as health care delivery, claims processing, and customer service. The value of process mining is discovery of processes as they actually occur without any transformation of the data. Specifically, process mining enables organizations to conduct an in-depth analysis of how current processes work, what should be automated, what can be automated, what benefits come from automation, and where process bottlenecks occur.
What if you could manage your event data by eliminating the need to pre-process data using complex ETL and data blending tools?
Download the white paper to explore this topic in-depth and see how ABBYY Timeline builds a virtual model of event logs using ELT, enabling data-driven process optimization.
1 Responsible Data Science: Using Event Data in a “People Friendly” Manner Prof. dr. ir. Wil van der Aalst, page 4