Information extraction for enhanced access to disease outbreak reports.

Area of Science:

• Information Science
• Public Health Informatics
• Computational Biology

Background:

• Traditional document search relies on individual terms, limiting efficiency for specialized collections.
• Accessing information within domain-specific document sets, like infectious disease outbreaks, presents unique challenges.

Purpose of the Study:

• To develop and evaluate Proteus-BIO, a novel system for enhanced document retrieval from infectious disease outbreak collections.
• To demonstrate the advantages of a structured, tabular database approach over conventional search methods.

Main Methods:

• Development of Proteus-BIO, incorporating a web crawler, an information extraction engine, and a database browser.
• Utilizing pattern and word class sets for information extraction, with automated discovery tools to streamline preparation.
• Comparative study assessing user efficiency using Proteus-BIO versus standard web search tools.

Main Results:

• Proteus-BIO successfully creates a searchable tabular database of infectious disease outbreaks, linking entries to source documents.
• Automated discovery tools significantly reduce the time required for preparing extraction patterns and word classes.
• A user study showed substantially higher document retrieval rates within a given time using Proteus-BIO compared to conventional search tools.

Conclusions:

• Proteus-BIO offers a powerful and efficient method for accessing and retrieving information from infectious disease outbreak report collections.
• The system's tabular database structure and automated information extraction facilitate more effective data discovery.
• This approach represents a significant advancement in managing and searching specialized scientific literature.