Last week TIBCO hosted the first Face2Face meeting of the decision modelling / management experts from FICO, IBM and Oracle representing one of the submission teams for the OMG Decision Model and Notation standard. Although we cannot divulge any details of the meeting results, suffice to say that progress was considered as “good” by all concerned… publication (to OMG) of the first draft of this and the other submissions is due in May 2012.

From an event processing perspective, current decision-based methodologies mostly focus on the “decision as a service” (DaaS?) type models and associated stateless executable artifacts - think TIBCO ActiveMatrix Decisions and TIBCO BusinessEvents Decision Manager, and their equivalents Oracle Business Rules, IBM Websphere Ilog JRules, and FICO Blaze Advisor. Here a decision is something that occurs at a point in time, based on data / information valid at the time of decision, with some explicit process required to revisit said decision.

However for event-based decisions, we can also consider decisions as a continuous operation. This is best illustrated by the “decision” made when some event pattern is identified: the pattern may be identified and/or disqualified multiple times for the same entity, based for example on complex and sophisticated business rules (that may themselves be changing over time). A typical mechanism for this might be inference rules (as seen in TIBCO BusinessEvents). I would expect later versions of DMN to be extended to cover rule-based and continuous-evaluation decisions, but DMN version 1 will likely cover the simpler “Decision as a Process Task” role (where “process task” could be a BPMN Business Rule Activity or a dynamic rule-driven activity). The differences in decision models for continuous versus decision-point based decisions is an interesting area for future discussion!

My thanks again to the participants of the F2F for a successful meeting.

I was going to blog on the presentations at the Decision Modeling Information Day at OMG, but suffice to say that James Taylor has already done a great job of this so I will just add a few comments from the CEP perspective…

1. CEP can be viewed just as “complex event” detection… but increasingly businesses want “complex  event” *processing*, where processing is not just event detection but also decision and reaction. Of course, the reaction can be a traditional orchestrated business process, as in a pattern like:

(a) detect fraud event (b) decide appropriate action (c) react with the Fraud Handling business process

2. Business modeling of decisions, as well as complex events, is an area that none of the traditional business modeling tools seem to be taking seriously. Yet both decision automation and CEP provide huge values to business operations.

3. Methodologists on the decision modeling side are finding receptive audiences in business. These include the KPI Decision Model, which was endorsed by Mark Pettit from Freddie Mac at the info day, the new BRS Question-Charts, and Alan Fish’s DRA. Adapting these for real-time, event-based decisions should be possible, as Larry Goldberg from KPI discussed in the meeting: the main difference with top-down analysis of decisions is that you may or may not be able to detect some of the desired business events!

4. There were some good questions on the role of analytics in, for example, driving CEP-based decisions, or in risk management for trading systems. For example, TIBCO Spotfire Data Miner can be used to generate rules for classification decisions inside a TIBCO BusinessEvents CEP system dealing with high volumes of trading events in a stateful manner.

5. The proposed DMN standard discussed at the end of the Information Day should of course equally apply to decisions in CEP as well as BPM.

OMG has posted the (latest) agenda for next month’s Decision Modeling information day - subject to change only if any other vendors sign up to show off their approaches to decision modeling. So far the agenda looks quite balanced:

• methodologies and notations: chiefly these are
  • Prof Jan Vanthienen (”Mr Decision Table”) and
  • Knowledge Partners Inc (authors of the book “The Decision Model”)
• vendors and tools: a good spead of capabilities here (updated!):
  • New Wisdom covering the repository of what we used to call “source rules” - the business texts for terms and facts as well as decisions
  • IBM Ilog, as the most successful BRMS tool around today, chiefly supporting the Websphere SOA model
  • TIBCO BusinessEvents Decision Manager, supporting the real-time decisioning and CEP market
  • OpenRules, and using Microsoft Office tools for decision management
  • Sapiens, covering their tooling for decision modelling

Keynote is being given by James Taylor (co-author of  “Smart Enough Systems”) who also promises to give his insights into what decision modeling is all about.

This is all by the way of introducing not only the ideas of decision modeling, but a proposed standard (DMN) in this space, to help enterprises like Wells Fargo and others who use multiple decision table, model and execution environments and want to improve efficiencies and re-use through standardised notations etc.

Should be an interesting session in an interesting week: the day after this the EPTS Virtual Symposium is taking place at the same location…

Analytics Magazine on OODA

While discussing the proposed OMG Decision Modelling standard with business rules specialist Ron Ross recently, Ron reminded me of the list of “roles” that “decisions” can provide:

Classification, Evaluation, Selection, Approval, Assessment, Assignment, Allocation, Diagnosis, Prediction

Curious as to what sources might refer to this list, I “googled” for these terms and one of the first relevant results was…

Threat evaluation and weapon assignment decision support: A review of the state of the art (2006-2007)

This is a fascinating paper that targets the military domain but nonetheless is very relevant to CEP (and indeed business) methodology:

1. “Threat evaluation” is also a business task: business threats are things like poor customer satisfaction, poorly managed product marketing, changing business conditions (locally or across the business), inability to predict business changes, regulatory compliance risks, and so on.
2. “Command and control” (a.k.a. business management) methodologies are also of interest: the paper refers to Boyd’s Orient-Observe-Decide-Act (OODA) cycle, Lawson’s Sense-Process-Compare-Decide-Act cycle, and a Monitor-Assess-Plan-Execute cycle.
   • Orient means setting up measurements, or event sources
   • Observe, Sense-Process-Compare, and Monitor means event (including “complex event”) detection, and all the associated processes for complex event processing.
   • Decide means making some decision based on our observed events.
   • Act and Plan-Execute mean behaviors that in turn will likely affect the environment and future events.
3. The paper refers to the JDL Data Fusion model which may well be familiar to some CEP folk: this model describes (with a minor translation from the table in the paper) the use of and relationships between:
   • event assessment
   • complex event or entity assessment
   • situation or state assessment
   • impact or cost-benefit assessment
   • performance or goal assessment.

I would therefore claim that the relationship between events and decisions seems not only very clear, but also well recognised.

DMN standard coverage (draft)

The OMG modelling standards group is creeping closer to releasing the Request For Proposals for DMN - the Decision Modelling standard. Decision-making of course is one of the primary goals of doing CEP - given an event pattern, what do I do about it? - so DMN is very much relevant to the CEP community. Recent newsworthy developments are:

• Following OMG Architecure Board comments to the effect that “surely you can do decision modelling in current UML Activity Diagrams etc”, the RFP team (TIBCO, IBM, K.U Leuven) have been working on clarifying the original RFP (which targeted Decision Modelling and Decision Management specialists) into a document suitable for a wider audience, better defined scope, and so on. We are working to get this for review and issuance at the March OMG meeting in Washington DC. The diagram to the right shows how the RFP is expecting the standard to look and how it should fit with some of the other modelling standards; of course there are plenty of other considerations too.
• The Decision Modelling Information Day is taking shape for Weds March 23rd 2011 in Washington DC, the day before the proposed virtual EPTS Symposium at the same hotel. This is targetting end-users of decision modelling and management tools, as well as tool vendors and consultants. The basic format is decision modelling approaches and use cases in the morning (including Jan Van Thienen on Decision Tables and KPI on their Decision Model approach) and associated decision tool implementations in the afternoon (including decision management and BRMS tool vendors). Indeed we want as many vendors as possible to present their approach - which as far as I know will be the first time this has been attempted! Could be interesting…

More updates to follow.