Build a business taxonomy in four steps
When it comes to using data across an enterprise, one of the most common pitfalls is not providing meaningful context to all data users. Managing business taxonomy within an enterprise is critical to support the day-to-day business operations within an organization because it ensures accurate and organized data that will be understood by all who need it.
When implemented successfully, a business taxonomy will become the foundation for content categorization and data relationships as well as a guideline that improves the speed of locating data while establishing policies on how it can be accessed or reused. Performing updates quickly is key to enabling an enterprise to respond immediately to compliance or emerging business-critical asks.
With our enhanced business governance artifacts updates, organizations have more ways to establish an enterprise-wide business taxonomy via IBM Watson Knowledge Catalog for IBM Cloud Pak for Data. The solution is also available as a fully managed service on IBM Cloud through IBM Cloud Pak for Data as a Service.
Four best practices for building a business taxonomy
Knowing where to begin can seem daunting. However, staying focused on four core activities and proven practices can help set up a successful journey to implementation.
1. Focus on a single high-value information area: Focus on a particular segment of the business that will drive the most significant impact. For instance, if GDPR and CCPA compliance is a high priority for your organization, begin with establishing terms and classifying assets related to personally identifiable information (PII).
The Policy Management features within IBM Watson Knowledge Catalog can enable data privacy and define data policies to describe how an organization should handle sensitive data. After creating policies, organizations can create Data Protection Rules to be automatically enforced and prevent unauthorized users from accessing sensitive data within a catalog. Organizations can also create Governance Rules to provide business descriptions of the required behavior or actions to implement a given governance policy. Business stakeholders can develop policies and governance rules based on governance subject areas that are important to the organization, such as information security, information privacy, or regulatory compliance.
Classifications, like tags, can classify and group assets based on your organization’s sensitivity or confidentiality level. Examples might be personally identifiable information, sensitive personal information, or assets deemed confidential. Organizations can also create a data protection rule in IBM Watson Knowledge Catalog to block users from accessing data assets based on its classification.
Furthermore, business stakeholders can use categories to organize governance artifacts and the users who can view and manage those artifacts.
Image 1: Include classifications in the conditions for data protection rules. For example, you can create a rule to deny access to certain users if the asset is classified as PII or confidential.
2. Concentrate on the meaning of business definitions: Use the language of your industry in the form of logical or business intelligence models to power existing terms and standards already set in place.
Business terms help define terms across the organization to have a unified understanding, allowing users to quickly find what they need and understand the tables and columns’ business context. For example, organizations can also use business terms to simplify the writing of data policy rules, using one business term instead of multiple different column names.
Organizations can also use business terms to link columns, assets, policies, and rules with the same type of data. For example, Reference Data allows customers to define a standard set of valid values for general lookup or define data classes. Organizations can also manage hierarchies and relationships in reference data and then relate a Reference Data Set to a business ontology.
3. Establish benefits and gain interest: Communicate to your organization to help them understand the advantages of having a single source of truth where all information is stored.
Within IBM Watson knowledge Catalog lies a central Knowledge Catalog that serves as a single source of truth for data engineers, data stewards, data scientists, and business analysts to gain self-service access to enterprise data that they can confidently trust and use. Organizations can drive self-service discovery and automate decision-making to evolve the business by providing and allowing access to a view of all information for those that need it.
Global Search within IBM Watson knowledge Catalog enables stakeholders to search across many projects, catalogs, assets, business terms, and other governance artifacts created within the organization. Collaboration features allow business units, users, and data owners to leave comments or assign a rating to an asset. Furthermore, data preparation tools like data refinery enable organizations to discover, cleanse, and transform data with built-in operations.
4. Develop and commit to milestones: Establish official milestones that your organization will commit to for implementing business categories, business terms, and correct assignment of user roles—and the data catalog process.
Each organization has unique needs where stakeholders in and out of IT need to add value to drive success.
The process flows for an organization may look something like the framework below:
Next steps
Try IBM Watson Knowledge Catalog with a two-month free trial of IBM Cloud Pak for Data on ANY cloud. New customers can receive 50% off the IBM Watson knowledge Catalog standard plan on the public cloud for six months.
Discover why IBM is recognized as a Leader in the 2020 Gartner Magic Quadrant for Data Integration Tools.
The post Build a business taxonomy in four steps appeared first on Journey to AI Blog.