<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Jenna Lau-Caruso on Medium]]></title>
        <description><![CDATA[Stories by Jenna Lau-Caruso on Medium]]></description>
        <link>https://medium.com/@jlau.caruso?source=rss-65453e61ffab------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*Y-nXdB0u_3tmrttPKTFO5Q.jpeg</url>
            <title>Stories by Jenna Lau-Caruso on Medium</title>
            <link>https://medium.com/@jlau.caruso?source=rss-65453e61ffab------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 07:49:57 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@jlau.caruso/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Your governance workflow is already ISO/IEC 11179-aligned: Mapping the registration lifecycle in…]]></title>
            <link>https://medium.com/@jlau.caruso/your-governance-workflow-is-already-iso-iec-11179-aligned-mapping-the-registration-lifecycle-in-619af68ad0ee?source=rss-65453e61ffab------2</link>
            <guid isPermaLink="false">https://medium.com/p/619af68ad0ee</guid>
            <category><![CDATA[iso-11179]]></category>
            <category><![CDATA[data-governance]]></category>
            <category><![CDATA[watsonx-data-intelligence]]></category>
            <category><![CDATA[iec-11179]]></category>
            <dc:creator><![CDATA[Jenna Lau-Caruso]]></dc:creator>
            <pubDate>Mon, 19 Oct 2026 20:25:07 GMT</pubDate>
            <atom:updated>2026-10-20T14:51:09.601Z</atom:updated>
            <content:encoded><![CDATA[<h3>Your governance workflow is already ISO/IEC 11179-aligned: Mapping the registration lifecycle in IBM watsonx.data intelligence</h3><p>Ask three teams to define “customer” and you will often get three different answers. Ask how mature each definition is, whether it is a rough draft, a reviewed proposal, or an approved enterprise standard, and the answers get even less consistent.</p><p>ISO/IEC 11179 is the international standard for metadata registries. It gives organizations a shared model for defining, naming, classifying, and registering data elements. ISO 11179-6 covers registration: the statuses an <em>administered item </em>moves through as it matures, and the roles that move it along.</p><p>If you already manage a business glossary in IBM watsonx.data intelligence, you likely have most of this in place. The built-in governance workflow maps to the ISO 11179-6 registration lifecycle.</p><p>In this post, we explore how the registration lifecycle and related ISO 11179 concepts map to IBM watsonx.data intelligence features. While the concepts can apply to all types of governance artifacts, in this walkthrough we focus mainly on business terms as an example.</p><h3><strong>Two kinds of status</strong></h3><p>ISO 11179-6 separates two ideas that are easy to conflate:</p><ul><li><strong>Registration status</strong> describes the quality and maturity a business term has reached, from Incomplete to Standard.</li><li><strong>Administrative status </strong>describes the step within the review and approval process that the business term is currently in.</li></ul><p>In watsonx.data intelligence, the workflow states you see (Draft, In Review, Published) are the administrative view. The ISO registration statuses sit on top of them. One workflow state can satisfy more than one ISO status.</p><h3><strong>The lifecycle at a glance</strong></h3><p>Here is the whole picture. Each box is a workflow state in watsonx.data intelligence. The pills inside show the ISO 11179-6 registration statuses covered by each state.</p><figure><img alt="How the watsonx.data intelligence workflow states map to ISO/IEC 11179-6 registration statuses" src="https://cdn-images-1.medium.com/max/747/1*D7Ugsk5qlTR3tK0dvU9w4Q.png" /><figcaption>Figure 1 — How the watsonx.data intelligence workflow states map to ISO/IEC 11179-6 statuses</figcaption></figure><h3><strong>Following one business term through the lifecycle</strong></h3><p>The easiest way to see the mapping is to follow a single business term from start to finish.</p><p><strong>Draft: <em>Incomplete</em><br></strong>An author creates a business term and saves it as a draft. At this point the term is only a proposal for others to see. No workflow has started and no version control is active. This is what ISO 11179-6 defines as <em>Incomplete</em>.</p><p><strong>Submit and Review: <em>Candidate, Recorded, Qualified</em><br></strong>When an author submits the term for publication, the governance workflow starts. This is the <em>Candidate</em> status.</p><p>ISO 11179-6 expects the stewardship organization to be recorded at this point. In watsonx.data intelligence, steward assignments live in the workflow configuration at the <strong>category level, </strong>so review tasks route automatically to the correct stewards for any given category.</p><p>The reviewer then checks the term. Are the name, definition, data class link, and category all present? Does the definition meet the organization’s quality rules? ISO calls these two checks <em>Recorded</em> and <em>Qualified</em>. In watsonx.data intelligence they are not shown as separate states, they happen together in the reviewer’s task. A term that fails either check is rejected and moved back to Draft.</p><p>This is where the standard’s definition rules (ISO 11179-4) come in. Reviewers can use workflow comments and rejections to enforce that definitions:</p><ul><li>Are written in the singular</li><li>State what the thing is, not what it isn’t</li><li>Are understandable without outside context</li><li>Do not use the term to define itself</li></ul><p><strong>Publish: <em>Standard</em><br></strong>When every approval step passes, the term is published as version 1.0. This is the <em>Standard</em> status. The term is now visible for metadata enrichment and discovery across the platform.</p><p><strong>Change and Retirement: <em>Preferred Standard, Superseded, Retired<br></em></strong>As terms change over time, the lifecycle covers changes as follows:</p><ul><li><em>Preferred Standard. </em>By default, the latest published version is the preferred one. If you need to designate a specific version explicitly, add a custom property such as <strong>Registration_Status = “Preferred Standard”</strong>.</li><li><em>Superseded</em>. When you publish a new version, the earlier versions are superseded automatically. Version history shows a link to the newer version, satisfying the ISO requirement that superseded items point to their successors.</li><li><em>Retired</em>. To retire a term, do not delete it. ISO 11179-6 expects retired items to stay in the registry for historical reference and audit. Handle retired terms with a custom property such as <strong>Registration_Status = “Retired”</strong>.</li></ul><h3><strong>What is built in? What requires configuration?</strong></h3><p>The table below puts all eight ISO registration statuses next to the watsonx.data intelligence mechanism that covers them.</p><figure><img alt="ISO 11179-6 registration statuses and the corresponding watsonx.data intelligence mechanism" src="https://cdn-images-1.medium.com/max/694/1*Kb77fTVcjX2sz277XwS6ag.png" /><figcaption>Figure 2. ISO 11179–6 registration statuses and the corresponding watsonx.data intelligence mechanism</figcaption></figure><p>The workflow supports the full lifecycle, and the intent of every status is met. A few statuses are enforced by reviewer practice or a custom property rather than a dedicated system state. Knowing which is which helps you decide what to document in your governance procedures.</p><h3><strong>Who does what? Mapping ISO roles to platform roles</strong></h3><p>ISO 11179-6 also defines the roles around the registry. Each one maps to a platform role and a category collaborator role within watsonx.data intelligence.</p><figure><img alt="ISO 11179–6 governance roles mapped to watsonx.data intelligence roles" src="https://cdn-images-1.medium.com/max/681/1*GPJAyvqqj9QhEGpMNxBw3Q.png" /><figcaption>Figure 3. ISO 11179-6 governance roles mapped to watsonx.data intelligence roles</figcaption></figure><p>Roles are assigned per category, so different parts of your glossary can have different stewards or committees. In practice, this is how most organizations already divide ownership.</p><h3><strong>Naming and identifying terms</strong></h3><p><strong>Standard names<br></strong>ISO 11179–5 defines naming conventions for metadata concepts. These conventions can be applied to business term names, and enforced through the review lifecycle, so that each name is unique and unambiguous.</p><p>Alternative names for use in legacy systems or local vocabulary can be handled using the business term’s <strong>alias</strong> field, or by creating secondary business terms as <strong>synonyms</strong> so that the standard name remains primary. Alternatively, custom properties such as <strong>label_en</strong>, <strong>label_de, label_fr</strong> can be used to capture multilingual names.</p><p><strong>Scoped identifiers<br></strong>A name tells people what a term is. An identifier tells systems which exact term and version they are referring to. ISO 11179–5 lets you form an identifier from three parts: the registration authority, a data identifier, and a version. In watsonx.data intelligence, immutable artifact IDs and versioning are used. If you operate across several organizations or business units and want a formal composite identifier, you can capture it with custom properties.</p><h3>Beyond the lifecycle: mapping the core concepts</h3><p>Registration is one of several areas ISO 11179 covers. The full set of watsonx.data intelligence governance artifacts support the ISO metamodel, naming, definitions and classification parts of the standard. Below is a short version of how the ISO 11179-3 core concepts map:</p><figure><img alt="Core ISO 11179 constructs mapped to watsonx.data intelligence features" src="https://cdn-images-1.medium.com/max/679/1*-XmKLlstCqodUJ_WPX7cvQ.png" /><figcaption>Figure 4. Core ISO 11179 constructs mapped to watsonx.data intelligence features</figcaption></figure><h3><strong>Where to go next?</strong></h3><p>If your governance program is already using IBM watsonx.data intelligence governance workflows, you are closer to ISO/IEC 11179 than you might think.</p><p>Ready to put it into practice? Explore the IBM watsonx.data intelligence documentation to start configuring your <a href="https://dataplatform.cloud.ibm.com/docs/content/wsj/governance/governance.html?context=df&amp;audience=wdp">governance artifacts</a> and tailoring your <a href="https://medium.com/medium.com/r?url=https%3A%2F%2Fdataplatform.cloud.ibm.com%2Fdocs%2Fcontent%2Fwsj%2Fgovernance%2Fworkflow.html%3Fcontext%3Ddf%26audience%3Dwdp">governance workflows</a> for ISO/IEC 11179 alignment.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=619af68ad0ee" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Ensuring consistent metadata in a multi-use data platform]]></title>
            <link>https://medium.com/@jlau.caruso/ensuring-consistent-metadata-in-a-multi-use-data-platform-d74f59887bd0?source=rss-65453e61ffab------2</link>
            <guid isPermaLink="false">https://medium.com/p/d74f59887bd0</guid>
            <category><![CDATA[data-catalog]]></category>
            <category><![CDATA[watson-knowledge-catalog]]></category>
            <category><![CDATA[data-intelligence]]></category>
            <category><![CDATA[data-governance]]></category>
            <category><![CDATA[ibm-knowledge-catalog]]></category>
            <dc:creator><![CDATA[Jenna Lau-Caruso]]></dc:creator>
            <pubDate>Fri, 24 Oct 2025 13:01:35 GMT</pubDate>
            <atom:updated>2025-10-24T13:08:33.623Z</atom:updated>
            <content:encoded><![CDATA[<p>In today’s data-driven enterprises, trust in metadata is not just nice-to-have, it’s foundational. When organizations rely on governance metadata catalogs to drive decisions, automate processes, and ensure compliance, inconsistencies can quickly erode confidence. Within platforms like <a href="https://www.ibm.com/products/watsonx-data-intelligence">IBM watsonx.data intelligence</a>, this metadata doesn’t sit idle, it actively powers downstream data products, AI applications and productivity tools.</p><p>Imagine discovering that the same table is represented in two different catalogs, each telling a different story about it’s business semantics, ownership or classification. Which one should you trust? This kind of ambiguity doesn’t just create confusion, it undermines the very purpose of data governance.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/787/1*1N3dF5aBKeL4jK1lejEtsg.png" /><figcaption>As the MORTGAGE_APPLICATION table moves between workspaces its classification can change in some catalogs but not others.</figcaption></figure><p>In this post, we’ll explore why metadata consistency is essential, not just for governance, but for enabling reliable, scalable collaboration across the enterprise — and how <a href="https://www.ibm.com/products/watsonx-data-intelligence">IBM watsonx.data intelligence</a> helps to ensure consistency of core metadata across the platform.</p><h3><strong>What makes two assets identical?</strong></h3><p>Asset metadata is a representation of a technical data asset, an abstraction that describes something which physically resides outside of the governance system. This could be a file stored in an S3 bucket, a table in a relational database, a view in a data warehouse, or a dataset in a lakehouse. The first step to ensuring consistency across catalogs is accurately identifying when two metadata entries refer to the same underlying asset.</p><p>This is easier said than done. In distributed or multi-node environments, asset identifiers can vary due to hostname changes, IP address shifts, or differing connection configurations. Without a reliable way to reconcile these variations, duplicate or conflicting metadata can proliferate, leading to confusion and inefficiencies across teams and tools.</p><p>To address these complexities, IBM watsonx.data intelligence introduced the concept of <a href="https://medium.com/@ankurverma0511/using-data-source-definitions-to-uniquely-identify-source-of-data-in-ibm-knowledge-catalog-07c38f640b57">data source definitions</a>. This construct allows administrators to explicitly define the endpoints and connection configurations that belong to the same logical data source, whether it is a database, lakehouse or cloud storage system. By doing so, the platform can assign a unique <strong>identity key</strong> to each asset, derived from it’s data source definition and connection path (eg. /database/schema/table). This identity key acts as a consistent fingerprint, enabling the system to recognize when two metadata entries refer to the same physical asset, even if they were ingested through different connections or reside in distributed environments. This foundational capability is critical for maintaining metadata integrity and avoiding duplication or fragmentation across catalogs.</p><h3><strong>Taking a layered approach to metadata</strong></h3><p>So, to avoid metadata discrepancies, can we just reject duplicates outright or maintain a single copy of each asset’s metadata in it’s entirety? If only it were that simple. In a multi-use data platform, different user personas operate in different workspaces, each with distinct expectations and responsibilities.</p><p>IBM watsonx.data intelligence is built on a unified catalog metadata repository which supports extensible asset types to power a wide spectrum of use cases. This flexibility is key to supporting the diverse needs across users of the platform. In this ecosystem, some metadata is highly context-specific, relevant only within a project, while other metadata is fundamental to the true representation of the asset.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Rcrfi_vw-jQJeNLpqyOmwg.png" /><figcaption>IBM’s multi-use data platform</figcaption></figure><p>A layered approach is employed to strike a balance between the two. In this model, certain properties of an asset can be designated as <strong>shared properties</strong>, meaning they belong to the asset’s core metadata and should remain consistent across the platform. This ensures that foundational metadata, such as schema definitions or data classifications, are preserved and trusted, while still allowing room for contextual metadata to evolve within collaborative workspaces.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/982/1*n6_RAHhG-fDFDiAtBCSasg.png" /><figcaption>Shared properties ensure consistency across workspaces, while non-shared properties allow for flexibility</figcaption></figure><p>Which properties are shared and which aren’t? Great question! The answer really comes down to purpose. Shared properties are those that describe the asset itself — things like schema, data classification, or asset owners. These are fundamental truths about the asset that should remain consistent no matter where or how the asset is used. On the other hand, non-shared properties reflect how the asset is being used within the platform. This includes things like how to connect to the asset, or which user last accessed the asset within a given workspace. By separating these layers, teams are enabled to work freely and iteratively, while still preserving a trusted, consistent view of the asset across the enterprise.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/813/1*vzC40rl-4vrh4xZnk-ehdw.png" /><figcaption>Base list of shared and non-shared properties. Custom properties can be declared as shared or non-shared according to their purpose.</figcaption></figure><h3><strong>How do shared properties impact each workspace?</strong></h3><p>Supporting a diverse range of user personas, tasks, and semantics requires a flexible workspace model — one that can accommodate both the collaborative, iterative nature of metadata work and the need for consistent, authoritative asset representations across the platform.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XPGZXYt7dFyapWBSPAwIaw.png" /><figcaption>Outline of shared property behavior across workspace types</figcaption></figure><p>Let’s walk through an example to see this layered metadata approach in action.</p><p>Suppose our governance catalog contains a table called <strong>MORTGAGE_APPLICATION</strong>. It’s been profiled to identify its data classes and quality, classified appropriately, and published as an authoritative asset. It’s also been packaged and shared across the organization as a data product.</p><p>Meanwhile, that same table is being used in two different projects, and is part of a production pipeline.</p><ul><li>In the first project, the <strong>Enrichment Project</strong>, the team is enhancing the metadata by assigning additional business terms and classifications. Here, the shared properties are temporarily moved into a <strong>draft state</strong>, allowing changes to be reviewed before they’re published back to the authoritative catalog.</li><li>In the second project, the <strong>Data Science Project</strong>, a group of data scientists is using the table in a notebook to train a machine learning model. They’re not modifying the metadata — they’re simply consuming the asset.</li><li>In the <strong>Production ETL Pipeline</strong> deployment space, all asset properties are frozen. Production jobs require explicit update to pick up changes in compliance with the organization’s change request process.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/660/1*P4Q1Fso18R0JTKq9dCwxdg.png" /><figcaption>Illustrative example of asset copies across workspaces on the data platform</figcaption></figure><p>As the review process wraps up in the <strong>Enrichment Project</strong>, the updated version of the <strong>MORTGAGE_APPLICATION</strong> table is published back to the governance catalog. Because the changes involved shared properties — those that define the asset itself — they’re immediately reflected across all instances of the asset on the platform. That means the <strong>Governance Catalog</strong>, the <strong>Data Science Project</strong>, and the <strong>Data Product Hub</strong> all now reference the updated, authoritative metadata. This ensures consistency without requiring manual updates, allowing teams to stay aligned while continuing to work independently and iteratively.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/649/1*heWKDSezCM8BHBVBryJ4IA.png" /><figcaption>Illustrative example of shared property behavior when asset updates are published to the governance catalog</figcaption></figure><p>Of course, this is just one example. In reality, the same asset might also appear in multiple governance catalogs, be part of several data products, or be actively used across a wide variety of projects. Thanks to the layered metadata model and shared properties, updates to core metadata stay consistent, no matter how widely the asset is used across the platform.</p><h3><strong>In Summary</strong></h3><p>In a modern, multi-use data platform, managing metadata effectively means embracing its complexity and nuance. IBM watsonx.data intelligence tackles this challenge through a layered metadata model, identity-based asset tracking, and flexible workspace constructs that support a wide range of user personas and tasks. By distinguishing between shared, authoritative metadata and context-specific annotations, the platform ensures consistency without sacrificing agility. This approach not only prevents duplication and confusion, but also empowers teams to collaborate confidently, ensure compliance, build high quality data products, and drive productivity across the enterprise.</p><h3><strong>Learn More</strong></h3><ul><li><a href="https://medium.com/@ankurverma0511/using-data-source-definitions-to-uniquely-identify-source-of-data-in-ibm-knowledge-catalog-07c38f640b57">Using Data source definitions to uniquely identify source of data</a></li><li><a href="https://www.ibm.com/docs/en/cloud-paks/cp-data/5.2.x?topic=catalogs-identical-data-assets">Identical Data Assets</a></li></ul><p><em>Note: The approach described above is available from IBM watsonx.data intelligence version 5.2, and is not applicable to adhoc catalogs which are not used for governance.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d74f59887bd0" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Effortlessly find the right Data with Semantic Search]]></title>
            <link>https://medium.com/@jlau.caruso/effortlessly-find-the-right-data-with-semantic-search-f4406abf222?source=rss-65453e61ffab------2</link>
            <guid isPermaLink="false">https://medium.com/p/f4406abf222</guid>
            <dc:creator><![CDATA[Jenna Lau-Caruso]]></dc:creator>
            <pubDate>Mon, 18 Jul 2022 13:12:02 GMT</pubDate>
            <atom:updated>2022-07-18T13:12:32.891Z</atom:updated>
            <content:encoded><![CDATA[<p>Learn how semantic search in Watson Knowledge Catalog 4.5 can help you find the data you need quickly and easily!</p><p>Check out the article I wrote with Lena Wolfe below:</p><p><a href="https://medium.com/@lwoolf_91808/effortlessly-find-the-right-data-with-semantic-search-cdb2bd9593ac">https://medium.com/@lwoolf_91808/effortlessly-find-the-right-data-with-semantic-search-cdb2bd9593ac</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f4406abf222" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>