Two events last month showed indicators of a convergence between the analytics world and CEP world.
Firstly Louis Bajuk-Yorgan from TIBCO Spotfire attended the Predictive Analytics World conference in San Francisco. He reported that:
Three key themes showed up multiple times throughout the talks-the growing importance of text mining, the application of net lift modeling to determine the real results of a marketing campaign (ignoring those people who would have bought anyway), and (most interesting to me) the importance of operationalizing predictive analytics.
In his opening keynote speech, Eric Siegel (the conference chair) saw the most important innovation in the field of Predictive Analytics focused on applying predictive analytics to operational decisions (as opposed to more established application areas such as customer churn & product recommendations). In a later talk, James Taylor of Decision Management Solutions (and co-author of the great book “Smart (Enough) Systems”), echoed Eric’s emphasis on operational results, encapsulated in the phrase “Action support, not just decision support.” James advised building an analytic platform that focused on the end game: the need to operationalize analytic decisions.
This is great validation for us, since operationalizing analytics is at the heart of TIBCO’s vision for its combined platform with Spotfire and S+ (as shown in products like Operations Analytics).
Then a week later Andreas Gerst from the TIBCO BusinessEvents team presented at cepconf in Munich, Germany. Andreas presented on CEP and Data Mining, and in particular how both these complement each other for advanced operational intelligence around customer management. Andreas used TIBCO BusinessEvents and TIBCO Spotfire Miner as his example technologies, mentioning techniques like PMML for moving from analytics to real-time event processing technologies.
FWIW I have noticed that people start to really consider statistics/analytics when they can see a clear path to using it.
So big and infrequent decisions become almost the low hanging fruit. Fortunately, they might be very juicy fruit as well. But big decisions make big targets for the data analyst, and infrequent decisions are even better targets.
Then there are those decisions that get made a million times a day, but individually have little impact. Sometimes it’s clear how to impact those decisions – like if the decisions are part of a web system, it’s just a matter of figuring out where to insert the decision.
And then there are the small decisions that aren’t so easy to control – well event-driven logic is one way to get control over them. And once there is some kind of control over the decisions, then there is a path to using analytics to make them better. And so it does not surprise me to hear that after starting on projects to get control over the small decisions, that there is increased interest in analytics. Because the event driven logic has provided a path for improving those decisions, including via analytics.
Hi Hans
What you say makes sense, and the decision hierarchy is something we’ve been looking at in the standards space (OMG Decision Model and Notation against various parts of the business motivation model). Probably I should post some more on that as its quite interesting…
The evolution of event-driven analytics is also reflected in the evolution of TIBCO BusinessEvents, if you think about it…
]
BE0-BE2: event pattern detection, rules and actions, some trending statistics
BE3: + explicit decision management
BE3+S+/Miner: + varying amounts of statistics and analytics
BEX: [...some other stuff TBA
Naturally, we in the CEP team take an “event-based view” of the analytics scene, whereas the analytics side of the house take a wider decision-improvement theme. Both sides agree on the need to evolve to provide “better operationalisation of the analytics” / “improving the performance of event pattern detection and discovery”. For sure the “low hanging fruit” for us will probably be exploiting Miner more in event processing – I think we discussed this concept of “event mining” earlier – and we hope to talk about some of the use cases over the coming months. One example: I heard of a manufacturing process application recently with an amazing ROI, yet the combination of event processing and statistics-derived knowledge would seem obvious to you…
I wonder if we’ll look back on this era with amazement that “to think we in IT used to try and do business optimisation apps etc without the statistician!”
I wouldn’t be surprised!
Overall, though, the combined approach of statistics in event processing will also require customer education and organisational challenges to be overcome – so its introduction will be a gradual process.
Cheers
Paul
For lots of companies the move to event-centric processing is what makes them see the power of operationalizing predictive analytics and automating decisions. It’s not the only use case, of course, but the complementarity is clear
JT
Hi James:
Interesting thought… I’m not sure though its the event-centric approach that encourages views of analytics per se – indeed in most organisations the analytics teams are still silo’d away from the operational IT guys.
Its more likely that as organizations see the need for (what we call) Operational Intelligence they start to understand that the practices and techniques required for this includes:
- knowledge-rich applications including embedding any analytics that can provide advantage
- event-centric views and responsiveness
It helps, of course, if they’ve read your book
Cheers