Metadata Management & Data Governance
Simply put, metadata is data that provides information about other data. In other words, they are the labels or tags that can be added to data to describe it and ultimately make it more functional, actionable, and understandable. The set of these activities, technologies, and policies aimed at organizing metadata is called metadata management. Successful metadata management is based on a solid data governance process and uses metadata standards that ensure consistency of processes along the different activities.
Key metadata management processes
As with any asset, it is good practise to take a process-oriented view on metadata management when designing the enviroment.
Metadata governance
While spreadsheets and relational databases are best used to support numerical data, much of an organization’s data, with an estimated 80% of business data across the globe, is rows upon rows of unstructured, text-based data. These large volumes of text require a great deal of manual time and effort to tag them with meaningful metadata and make them usable.
Metadata standards
Machine learning and vocabulary-based classification and entity extraction can be used to find metadata that best describes the data and conforms to conventions set by management. This ensures that tagging is done within the defined range of values.
Metadata creation
Ideally, the creation of metadata should be done directly by the data authors themselves, and supported by digital assistants. It has been shown that the integration of metadata management systems as an auxiliary system directly into the data creation systems ensures this best.
Metadata discovery
Machine learning and vocabulary-based classification and entity extraction can be used to find metadata that best describes the data and conforms to conventions set by management. This ensures that tagging is done within the defined range of values.
Metadata quality
Especially with fast-growing knowledge models, it can happen that relations are set incorrectly or inconsistencies creep in. Quality checks ensure that metadata guidelines are applied correctly and that specified schemas are adhered to.
Metadata storage
Knowledge model and instance data are stored as No-SQL in tripple stores and are thus directly addressable and linkable by SPARQL.
The Semantic Web Company—vendor of PoolParty—has been named a Visionary in the Gartner Magic Quadrant for Metadata Management Solutions. Among the strengths of the PoolParty solution, Gartner emphasizes the combination of first and second-generation machine learning, the beneficial combination of knowledge graphs and NLP, and the combination of text mining and semantic classification to develop a comprehensive suite that supports multiple scenarios.
How rich metadata changes the way you deal with data
Praised by Gartner Peer Insights for metadata, the PoolParty Semantic Suite combines Knowledge Graphs, Text Mining, Natural Language Processing and Machine Learning to fundamentally change the way organizations deal with data. With PoolParty, companies can build a semantic layer that, when applied to their information architecture, enables them to automatically analyze and create rich metadata. With semantic metadata, they can bridge data silos and gain a unified view of structured and unstructured data.
Data compliance
Data compliance is one of the most demanded use cases for metadata management. Besides helping comply with regulatory and business requirements, metadata management helps assess the impact of a change within a data source. It also supports accountability for the terms and definitions of a business glossary to lead organizations towards the development of a compliance data model.
Process control
In a practical scenario, metadata management can help determine the next steps in a workflow or business process. Rules based on metadata can help to find the best possible interactions for a data object, but also to guide and prioritize interaction options among each other. We can speak here of metadata-guided business logic.
Strong analytics
Monitoring performance and making valuable decisions based on data should be at the forefront of any organization. Metadata management supports building a data catalog for the analytical uses of data across an organization. As a result, organizations avoid confusion and misinterpretation of information, which strengthens their ability to make educated business decisions.
Highlighting risks
If you include additional vectors in the metadata, you can include information about risk potentials, assessment frameworks, and classifications in the description of data. You get quality metadata that gives you sentiment and prospects, risk analysis, and alerts for your BI.
Useful Resources
Success Story:
Consistent metadata for international cooperation with REEEP
Webinar:
Watch “Data-driven knowledge management with knowledge graphs” by Heather Hedden and Fredric Landqvist.
Webinar:
Boost Customer Experience with Auto-Tagging & Metadata