A Recommender System to Help Refining Clinical Research Studies.

Area of Science:

• Medical Informatics
• Information Retrieval
• Bibliometrics

Background:

• Refining research questions in medical studies relies heavily on existing background knowledge.
• Prior publications significantly influence study stages, particularly cohort definition.
• Information from cohort selection tools can enhance and expedite research but is often not indexed by search engines.

Purpose of the Study:

• To present a novel methodology for identifying relevant scientific publications.
• To address the challenge of discovering information not typically captured by standard search engines.
• To aid researchers in refining research questions and defining study cohorts.

Main Methods:

• A hybrid approach combining content-based retrieval and text annotation techniques.
• Utilizing these techniques to identify publications relevant to specific research questions.
• Linking identified publications to selected data sources.

Main Results:

• The methodology effectively identifies relevant scientific literature.
• It enhances the ability to find information beyond traditional search engine capabilities.
• Facilitates enrichment and acceleration of research processes, especially in cohort definition.

Conclusions:

• The proposed methodology improves the discovery of scientific publications for medical research.
• It offers a valuable tool for researchers, particularly in the early stages of study design.
• Enhances the integration of diverse information sources for more robust research question refinement.