Mapping the Studies Behind Every Filing: Extending a Pharmaceutical Taxonomy for Regulatory Search.

About the Client

Company Z is a global pharmaceutical manufacturer that sells its products in more than 140 countries. Company Z’s manufacturing division oversees the formulation, packaging, and distribution of its global product portfolio. Its Knowledge Management Center of Excellence has partnered with Iknow on multiple engagements to build and continuously expand the enterprise taxonomy that powers search across Company Z’s SharePoint repositories.

Industry Context

Bringing a new drug to market and keeping it there requires an extensive body of Chemistry, Manufacturing, and Controls (CMC) documentation — stability studies, process characterization and validation studies, impurity profiling, and more — that regulators such as the FDA require as part of INDs, NDAs, BLAs, and their international equivalents. Stability data packages are often the most extensive nonclinical sections of a regulatory submission, and any planned change to a raw material, manufacturing step, or packaging system can trigger the need for additional studies and filings. For a manufacturer the size of Company Z, quickly locating the specific prior study relevant to a new filing question, rather than starting from scratch or relying on institutional memory, can materially affect how quickly a regulatory submission comes together.

Current Situation

In prior engagements, Iknow built Company Z’s core enterprise taxonomy — more than 1,840 terms across a four-level hierarchy and 10 facets — with autoclassification rules deployed across Company Z’s SharePoint repositories, iterating further based on testing, feedback, and analysis of content such as manufacturing deviation reports. Company Z’s KM Center of Excellence wanted to further extend the taxonomy specifically for the commercialization process, where a large number of specific studies are required for regulatory approval, and details of those studies were hard to locate quickly. It also asked Iknow to explore cross-topic relationships that could improve both search and autoclassification.

• Regulatory study content buried in a taxonomy not built for it. Company Z’s existing taxonomy lacked sufficient coverage of the specific studies required at each commercialization step, making it hard to quickly find relevant content.
• High-stakes content with no room for search failure. Regulatory filings depend on accurately locating prior stability, process characterization, and other CMC studies; slow or incomplete searches directly affect filing timelines.
• Taxonomy structure that had drifted from real regulatory workflows. The commercialization-step taxonomy needed to be rearranged to better align with how study sequences actually unfold during the regulatory process.
• Untapped value in relationships between topics. The taxonomy’s terms existed largely as an unconnected list, without a systematic way to surface related concepts across commercialization steps, document types, and manufacturing steps and issues.

• • Extend Company Z’s taxonomy to better cover the studies required for regulatory approval during commercialization.
• • Add terms and synonyms and adjust autoclassification rules to ensure relevant regulatory study content is tagged more accurately.
• • Explore and pilot additional cross-topic relationships in the taxonomy to improve search, browsing, and autoclassification.

How Iknow Structured the Work

Iknow structured the engagement into four phases: user interviews to ground the work in real needs, taxonomy extension, implementation and testing, and a pilot project to explore additional cross-topic relationships.

Key Activities & Decisions

• User interviews. Iknow interviewed Company Z’s search users across multiple departments and locations to understand how studies are currently filed and classified and how the process might be improved. Iknow decided to base the taxonomy extension on Company Z’s study maps — documents listing all studies required at each step of the regulatory process for a given product type.
• Taxonomy extension. Iknow used study-map terminology to add concepts to Company Z’s main taxonomy, adding about 100 new terms, synonyms, and alternative labels, and reordering the existing commercialization-step taxonomy to better align with study sequences. Iknow developed the additions first in Excel for quick user review, then implemented them in the Semaphore Ontology Editor.
• Implementation and testing. Iknow tested the new taxonomy in the autoclassification engine against about 50 documents, confirming that no changes were required to the existing autoclassification rules.
• Cross-topic relationship pilot. Iknow reviewed the extended taxonomy to identify additional cross-topic relationships beyond those already in the model, adding further relationships and relationship descriptions, especially between commercialization steps and corresponding document types, manufacturing steps, and manufacturing issues.

Stakeholders & Collaboration

Iknow served as prime contractor, interviewing Company Z’s search users across multiple departments and locations and continuing its long-running taxonomy partnership with Company Z’s Knowledge Management Center of Excellence.

Extending a Mature, Live Taxonomy Without Disrupting It

Company Z’s taxonomy was already deployed and actively powering autoclassification across the firm’s SharePoint repositories, so any extension risked degrading existing search results if done carelessly. Iknow addressed this by grounding every new term directly in Company Z’s study-map documentation rather than generic industry vocabulary, and by staging changes in Excel for user review before implementing them in the Semaphore Ontology Editor, validating additions before touching the live autoclassification model.

Proving Cross-Topic Relationships Were Worth the Investment

Relating every term in a 1,840-term taxonomy to every other relevant term would have been an open-ended, difficult-to-justify undertaking. Iknow addressed this by scoping the relationship work as an explicit pilot focused on the highest-value connections — commercialization steps to document types, manufacturing steps, and manufacturing issues — rather than attempting to relate every term in the taxonomy at once.

Quantitative Outcomes

• New terms added: approximately 100 new terms, synonyms, and alternative labels.
• Testing: roughly 50 documents were tested in the autoclassification engine, with no existing rule changes required.

Qualitative Outcomes

The expanded taxonomy delivered strong classification results, significantly improving the tagging of commercialization documents and giving Company Z’s scientists and engineers a faster path to the specific stability, process characterization, and other CMC studies needed for regulatory filings. The cross-topic relationship pilot added a new dimension to search and browsing, allowing a single tagged term to point users toward related concepts they might not have considered searching for directly, while also providing the autoclassification engine with additional evidence to tag related terms. The work was completed on schedule, and Company Z planned to further develop these capabilities in the mid-2020s, building on the taxonomy foundation established across Iknow’s multiple engagements.

Timeline to Impact

Iknow delivered all four phases within the five-month contract period, with results validated against real Company Z content before final delivery.

Core Skills

• Taxonomy extension & governance
• Regulatory content classification
• Semantic relationship modeling
• User-centered requirements gathering

Methods & Frameworks

• Structured user interviews
• Study-map-grounded taxonomy design
• Staged Excel-to-Ontology-Editor implementation
• Autoclassification testing and validation

Technologies & Tools

• Progress Semaphore Ontology Editor
• SharePoint enterprise search
• Autoclassification engine