OCMiner is a robust text processing system optimized for scientific publications, particularly in the life sciences. It recognizes and annotates entities such as chemical compounds, proteins, and diseases, mapping these to domain-specific ontologies to extract semantic relations. Using modular pipelines, OCMiner handles complex text sources and recognizes terms with varied structures, from chemical formulas to abbreviations. The system achieved high precision in chemical named entity recognition in the BioCreative IV challenge, highlighting its speed and accuracy. By indexing recognized entities and relationships, OCMiner enables intuitive exploration and retrieval of interconnected scientific data.