[Big data, medical language and biomedical terminology systems]

Area of Science:

• Biomedical Informatics
• Natural Language Processing
• Big Data Analytics

Background:

• Healthcare and biomedical research heavily rely on structured terminologies like thesauri and ontologies.
• However, natural language in electronic medical records and scientific publications remains the primary information carrier.
• Bridging this gap is crucial for advancing data-driven biomedical insights.

Purpose of the Study:

• To explore the hypothesis that structuring natural language into normalized information facilitates big data analytics in biomedicine.
• To investigate the potential of combining natural language technologies, semantic resources, and big data methods for knowledge extraction.

Main Methods:

• Utilizing evolving computerized human language technologies for annotating biomedical narratives with standardized codes.
• Employing big data methods to support the creation and maintenance of linguistic and terminological resources.
• Developing methods for learning hierarchical relationships, grouping synonyms into concepts, and disambiguating homonyms.

Main Results:

• The study posits that abstracting natural language into structured, semantically normalized data enhances statistical analysis.
• Big data methods show potential in supporting the labor-intensive creation and maintenance of essential linguistic and terminological resources.
• Examples of big data applications include identifying hierarchical relationships, concept grouping, and homonym disambiguation.

Conclusions:

• The integration of natural language technologies, semantic resources, and big data analytics holds significant promise for biomedical knowledge discovery.
• While direct evidence is still emerging, this combined approach is expected to unlock new insights from vast amounts of textual data.
• This synergy can improve the utilization of clinical and research data for improved healthcare and scientific advancement.