Table of contents for Live from DAMA 2008
- Live from DAMA – Yin and Yang of Process and Data
- Live from DAMA – What do they mean, BUSINESS Metadata?
- Live from DAMA – Business rules, decision management and smarter systems
- Live from DAMA – A Reference Architecture for Integrating an Active Data Warehouse into the Real-Time Enterprise
- Live from DAMA – Re-Thinking Search in a Web 2.0 World
- Live from DAMA – Deploying Semantic Web Solutions: I’ve Built an Ontology, Now What?
- Live from DAMA – The Business Drivers Behind Creating an Enterprise Data Architecture in the Gaming Industry
- Live from DAMA – Metadata enabled Business/IT Integration
- Live from DAMA – IBM Metadata Strategy – An Information Management Perspective
- Live from DAMA – Impact of Service Oriented Architecture on Data Modeling: A Case Study
- Live from DAMA – Case Study: Implementing a Securities Master Using Flexible Data Models at Lord Abbett
- Live from DAMA – Naked Without My Data
Neil and I are attending the DAMA conference this week and I will be blogging from some sessions. First one (after a fairly long set of announcements) is the Tuesday morning keynote, Michael Blechar of Gartner on The Yin & Yang of Process and Data: Which Will Be King of the Next Generation of Applications? Today, lots of business process improvement is being done with Business Process Management Systems based on a Service-Oriented Architecture. While this is effective, it can tend to lose sight of data. Michael feels that Process and Data are a Yin and Yang – they are intertwined and both essential.
Information Architecture is critically important to SOA/BPM initiatives. Business Process agility requires information understanding. Users will increasingly use a composition platform for process orchestration driven by dashboards and “expert system” components – next generation BI solutions. This is supported by a business services registry that is managed and that controls access to custom applications, packages and external services. These manage business content and business rules. All of this drives information management requirements.
60% of process improvement is human to human, second category is orchestrated workflows using process and information services (and decision services, to me), third is content-driven solutions that are very information intensive.
SOA decouples the user interface from the services and from data – services use data but are often reused between process and applications which raises issues around data ownership, security, change management and more. in addition, BPM approaches use different models that don’t necessarily support the kind of data analysis typically needed for information management and services can obscure where data comes from. SOA exposes data issues to more people, places and processes and you must figure out how to manage information in this new environment. Michael sees a layer of information services handling things like metadata, business rules etc. Best Practices in information management to support and enable SOA/BPM:
- Information management is an integral part of Enterprise Architecture – Business Architecture, Information Architecture, Technical Architecture coming together in a Solution Architecture.
- Align business information and technical architectures at various levels/stages – conceptual, implementation or whatever. Need models across the different architectures at each level so they can be discussed as a set.
- Promote reuse through architectural patterns and Master Data Management. Unless data services are designed and metadata managed then reuse is going to be difficult or problematic. This might be a data warehouse but more likely to involve some kind of master data hub that supports real-time federation of data from multiple sources.
- Wrap selected legacy data to improve transparency and productivity
- Put process-centric ‘data in context’ rules in separate services. Take logic out of stored procedures and put it into proper services. Some services will be basic CRUD services – data centric – but others should handle the rules that go with the data in context. I would call these Decision Services and Michael talks about these as a critical part of the “future data architecture”.
- Promote understanding, compliance and reuse of information assets. This involves thinking about modeling approaches and methodologies as well as metadata registries and repositories. Increasingly this involves a federated solution which is not quite there yet.
- Manage and share information assets via Meta Models – identify, rationalize and share information assets so can see how the information infrastructure is being impacted
- Build an Enterprise Information Management reference model to support BPM and SOA. This needs a layer of metadata and semantics to support data services.
- Develop an EIM road map to enable to the information architecture, then execute it.
- Balance process-driven solutions with content-driven ones. Service-oriented will increasingly be balanced with information-centric (decision-centric) as part of the service portfolio. Michael sees that enterprise mashups become more important and bring together content services and move from BI to Business Activity Monitoring.
- Coordinate all aspects of enterprise architecture
- Wrapper legacy applications and data, use Master Data Management
- Buy only service-oriented solutions
- Use data in context services for rules
- Use models and metadata repositories for understanding, compliance and governance
- Implement an Enterprise Information Management reference model and program
- Start adding content-driven improvements to process-centric improvements
Nice to see lots of focus on business rules in Michael’s presentation. I think that Decision Services (wiki), by bringing rules and analytics to the service layer, can address many of his items. Decision Services can be composited into applications by business users to make their processes “smarter”, especially when these processes are delivered through automated channels or by low level front line staff. These decision services combine data, insight from data and business semantics into reusable services.