About

How does MDPI ROR Entity Search solve internal challenges?

MDPI ROR Entity Search was created to solve a concrete internal problem by turning messy, free-text affiliation strings into standardized organization identifiers. With the high publication volume at MDPI, the need to reliably link authors and institutions has become critical for accurate analytics, reporting, and process automation across the publication lifecycle.

Affiliations in manuscripts are highly inconsistent: the same institution can appear in dozens of formats, languages, and abbreviations. This makes it difficult to answer simple questions like, “How many papers did we publish with Institution X last year?”, support institutional agreements (IOAP), and design internal workflows and dashboards that rely on clean organizational data. Existing tools and manual curation could not scale to MDPI’s volume, nor provide the precision and context-awareness we needed.

To address this, we built an AI-powered ROR Entity Search API: a centralized service that disambiguates affiliation strings and returns standardized ROR IDs with confidence scores. The service combines named-entity recognition (NER) to extract organization names from free text, and semantic search over vector embeddings to match these entities against the Research Organization Registry, even when names are incomplete, translated, or misspelled. It supports bulk processing to handle MDPI’s throughput and is refreshed frequently with the latest ROR data to keep mappings accurate and current.

By integrating this service into our internal systems, we can:

• Produce consistent institution-level analytics and dashboards
• Better support institutional agreements and reporting obligations
• Improve metadata quality across MDPI platforms
• Lay the foundation for more advanced, organization-aware tools and workflows

In short, MDPI's AI-powered ROR Entity Search API operationalizes affiliation data at scale, enabling MDPI to move from free-text affiliations to robust, standardized organizational intelligence.