About DeCS Finder AI
DeCS Finder AI is a tool developed by BIREME/PAHO/WHO to support the organization and retrieval of health information through the automatic discovery of descriptors from the thesaurus DeCS – Health Sciences Descriptors.
DeCS, developed by BIREME since 1986, is a controlled vocabulary that expands the MeSH (Medical Subject Headings) of the NLM, incorporating terms relevant to the reality of countries in Latin America and the Caribbean, especially in public health, collective health, and health surveillance. It promotes semantic standardization and interoperability among information systems.
History
DeCS Finder was launched in 2021, initially using only a lexical algorithm of Natural Language Processing (NLP) called DeCS Highlighter.
In 2024, the beta version of DeCS Finder with Artificial Intelligence (AI) was released, incorporating more advanced text analysis models for descriptor recommendation based on the LILACS Methodology.
How It Works
DeCS Finder AI works through the automatic analysis of texts using Natural Language Processing (NLP) techniques and AI models.
When entering a text – title, abstract, or full text, the system processes the content, identifies concepts and topics, and suggests descriptors from DeCS that represent the main subjects.
The AI model was trained based on data from manual indexing of journal articles indexed in the LILACS database, carried out by the VHL Network and collaborating centers, following the LILACS Methodology.
The system combines statistical, semantic, and lexical algorithms to offer relevant and accurate recommendations aligned with the analyzed content.
Based on AI and NLP, DeCS Finder AI supports the process of assisted indexing, contributing to document description and improving information retrieval in databases, institutional repositories, journal portals, and other information environments.
Target Audience
The tool is designed to support information professionals, librarians, researchers, and institutions involved in managing and disseminating scientific knowledge in health.
Additionally, DeCS Finder AI can be used by researchers and authors as a support tool in selecting descriptors when submitting scientific texts to journals and repositories, contributing to the accurate thematic representation of their works.
Its use supports semantic standardization, interoperability among information systems, and more precise and efficient content retrieval.
⚠️ Important: DeCS Finder AI is a support tool for indexers and does not replace human analysis.
Descriptor suggestions should be carefully reviewed and validated based on a thorough reading of the full text to ensure correct and accurate term assignment.
Main Features
- Automatic discovery of DeCS descriptors for texts in Portuguese, Spanish, and English;
- Intuitive and user-friendly interface;
- Processes titles, abstracts, or full texts;
- Exports results for use in indexing, data curation, cataloging, and text submission activities.
Benefits
- Reduces time and manual effort in descriptor discovery;
- Optimizes indexing and curation processes of scientific content;
- Improves the quality and precision of information retrieval;
- Enhances the visibility and accessibility of scientific content by using an internationally recognized controlled vocabulary;
- Promotes interoperability among different health information systems.
DeCS Finder AI strengthens the commitment of BIREME/PAHO/WHO to improving health information management in the Americas.
Its creation reflects the transformation of information products and services for the digital environment, incorporating emerging technologies such as artificial intelligence to make processes more efficient, scalable, and aligned with the demands of the digital world.
