Automated MeSH indexing of the World-Wide Web

Area of Science:

• Biomedical Informatics
• Information Science
• Computer Science

Background:

• Effective information retrieval is crucial in the vast biomedical domain.
• Manual indexing of web-based biomedical documents is time-consuming and inconsistent.
• Networked discovery requires standardized methods for organizing research data.

Purpose of the Study:

• To design and implement an automated system for assigning Medical Subject Headings (MeSH) to biomedical documents.
• To facilitate networked discovery and enhance information retrieval from World-Wide Web sources.
• To develop a prototype for an automated indexing tool, the 'Web-MeSH Medibot.'

Main Methods:

• Developing algorithms for automatic MeSH term assignment to web-retrieved documents.
• Implementing prototype systems to test the feasibility of automated biomedical document indexing.
• Utilizing natural language processing and machine learning techniques for term identification.

Main Results:

• Prototype implementations demonstrate significant promise in accurately assigning MeSH terms.
• The system shows potential for automating the indexing process for biomedical literature.
• Initial results suggest feasibility for a fully automated indexing solution.

Conclusions:

• Automated MeSH assignment is a viable approach to improve biomedical information retrieval.
• The 'Web-MeSH Medibot' concept offers a pathway towards efficient, large-scale biomedical data indexing.
• Further development is warranted to create a completely automated indexing tool.