Datafication is an unstoppable trend for the coming years. Its bottom line: data is as an asset instead of a by-product of processes and applications.

Companies like Facebook have caused an emphasis on the value of data – both commercially and strategically. New technologies allow to build applications instantly to support new business models and customer needs, what nowadays causes data to outlive applications. Customer records may be around for decades and are growing exponential because of the trend of turning many aspects of our life into computerised data. The applications, functions and processes that use this data may be plenty and built when needed.

Just like other assets, data represents value: Either in dashboards that show opportunities to cut operational cost; as intelligence that depict market opportunities; as predictive analytics that reduce customer churn; or as simply as data-sets that may be sold for commercial use. The market is beginning to recognize the importance of data, and Data Management gains importance.

Data Management

Data Management is not new: this discipline has been around for ages. Already in 1989 an international community founded DAMA (the Data Management Organization). This association proposes an exhaustive set of practices and models to provide guidance in this field. The latest update of the DAMA/PM-Bok handbook is only from 2017. For most data related issues it provides organizations with loads of instruments to manage their data. It addresses subjects like governance, data quality, data architecture, master data management, data modelling and meta data. But also BI and data warehousing, storage, data security, data integration and document and content management.

DAMA functions model

DAMA offers several knowledge products and models that look together like an exhaustive set. Turning them into practice however, they still seem to be lacking certain recent innovations in the area of Big Data, Artificial Intelligence, Realtime data processing and API management. Some of those are just slightly represented in the most important DAMA models. To make DAMA applicable in an innovative environment we at SynTouch extended the DAMA models. Take for example the DAMA Data Functions Model. Currently DAMA has defined it as follows:

To support the mentioned innovations we have added new knowledge products to this model :

• Artificial Intelligence / Deep Learning: AI is a new field that involves new challenges. AI may determine its own data sources, create (meta) data, and may (in the future) may need to comply to legislation to track/trace actions back through a path of data and models that caused the decisions to be made.
• Data presentation and visualization: As data size and complexity increases and the usage of social media is growing, the need for effective communication (e.g. through visualization) of data increases too. This requires new approaches, technologies and methods that require their own topic in DAMA.
• Data-API’s: As data is recognized as an asset it will be of increasingly importance in the Data-Supply-Chain. The obvious transport-means is internet. Data-API-s are a flexible way to distribute data on demand. It combines well with pay-per-use. Setting up Data-API-s in a correct way is a specific knowledge area. It involves registration, well-defined data message models, metering, granularity, security, anonymization and integration to disclose the backend-systems.
• Big Data storage: Besides the structured data in the Data Warehouse it is common practice to store semi-structured or unstructured data in a data-lake, and (often after processing) in specific no-SQL databases. This requires with a specific set of storage functions different from the traditional ones.
• Blockchain: Blockchain is a specific technology for distributed storage of transaction data, so non-repudiation is guaranteed. The participants are responsible for validating the transactions. Hence no central party is required anymore, which causes that blockchain is considered as a big game changer for the next years.
• Data streaming and Event Processing: In order to support AI and machine learning big data needs to be processed in real-time. This requires specific solutions that exceed data integration functions. Think of streaming data ingress and egress, streaming processing (think of tools like Spark, Storm, Fluke) and in-memory databases.
• Data Acquisition: As data is an asset it may be acquired directly. It involves a number of purchase related functions like data specification, vendor selection, valuation, intake and possible aftersales.

The Systems / Applications-box of the DAMA diagram is extended with processes, Because an increasingly amount of data will be produced or consumed outside applications in Business Process Management Workflows through worklist and task-screens.

The Systems / Applications-box of the DAMA diagram is extended with processes, Because an increasingly amount of data will be produced or consumed outside applications in Business Process Management Workflows through worklist and task-screens. Data Warehouse is extended with Big Data storage, in order to include also semi and unstructured data for intelligence purposes.

This model provides a more complete picture of actual Data Functions. A next blog will elaborate on other DAMA products.