Behavior Mining vs Process Mining
Behavior mining aims to discover, monitor, and improve end-user behavior (customers & employees). It does that by extracting knowledge from transaction and event logs readily available in today’s information systems. Process Mining and market abuse systems seek to uncover insight hidden in FIX drop-copies or system logs, whereas Behavior Mining seeks to uncover insight from the behavior of the user. Process Mining tools are typically limited to single applications and Behavior Mining allows tracking application usage across a broad range of internal applications.
Top 5 Use Cases of Behavior Mining
1. RISK AND COMPLIANCE
From a Compliance and Risk Management perspective, capturing all user data is useful to investigate any incident, regardless of the time passed since the event occurred. Does the employee recall why he/she deviated from the prescribed path and how they sought to understand the risks involved?
2. OPERATIONAL EFFICIENCY AND PROCESS AUTOMATION
The business operations process is the amalgamation of multiple customer journeys, brought together in a system. If we were to document an application process or data flow using UML, the ‘user touch-points’, are invariably represented as the steps in the process diagram. Every single step has an input, an output, with some human intervention, or action in the middle. Behavior Mining gives us insight into all these interactions, and importantly which interactions can be improved. For example, in a recent customer engagement, which deployed Behavior Mining toolkit, it was discovered that users were making 25 “copy and pastes” to complete a single stage of a process.
3. APPLICATION MODERNIZATION
Consider the ongoing cycle of investment to modernize the internal application estate; there is a significant advantage to start with a thorough understanding of how the business uses applications today. Without this insight, firms start with management’s opinion of how they believe the application is being used. A more scientific approach that starts with a real understanding of who is using the application, for what purpose, is more suitable to get the best outcome of an application refresh. This use case could also leverage historical data (rather than real-time), and thus the infrastructure and processing costs of such a solution would be relatively modest.
4. WORKFORCE OPTIMIZATION AND COACHING
Behavior Mining as a discipline is useful for workforce optimization provided it isn’t used to micro-manage employees. Both executives and their employees are quick to highlight that the technology can be used for “tracking” or “policing” individual employees, but in reality, this is an edge case. In financial services, a certain degree of tracking is not only common but often mandated by the regulators. Thus resistance to monitoring technologies is likely to be much less within financial services than perhaps other industries.
5. REDUCING TERMINAL LICENSING COSTS
Usage metrics of costly applications, like Bloomberg terminal or Oracle licenses, are often limited to login counts per month. Various tools have been created over the years to interpret usage (to differing degrees of success). Those involved in such license reduction exercises will know that the results are often unreliable, and don’t necessarily provide enough evidence to justify cancellations.
Implementation Challenges
The systems we seek to analyze have often been designed without any thought being given to the need to mine their user’s behavior. This often results in sparse output, challenges in drawing correlations between events, and indeed at times inconsistent logging of the events in the first place. This is normal as the systems in question might not have been designed to record users’ behavior, rather the focus is to record system errors and transactions so that IT teams are able to have a view of what’s gone wrong with their systems.
The Solution
Currently, Glue42 is the only commercial offering that enables operational, compliance, and risk teams to capture and then visualize users’ behavior as they navigate within and between applications. The platform allows you to:
- Monitor and record all user behavior, in real-time, across 10s of thousands of desktops
- Mine the data looking for application usage, process anomalies, opportunities for improvement, compliance, etc.
- Feedback evidence of usage into operational teams, training departments and IT change teams
For a personal demonstration you can book a 1-on-1 demo.
Want to Learn More About Behavior Mining?
Download latest whitepaper by financial consultancy Rocketfin.
