The Effectiveness of the Text Mining and Similar Document Search System for Evidence-based Guideline Development

Abstract

 Objective: The objective of this study is to evaluate the effectiveness of using the text mining and other document search systems for the development of evidenced-based clinical practice guideline.
 Methods: We arranged document databases for searches, exercises and adequate documents to evaluate these systems. We used the precision and recall rate as evaluation criteria.
 Results: By using the text mining system, we achieved a high precision on the clinical question for the guideline, while keeping a high recall rate. By using the similar document search systems, we achieved a good result with the dimension reduced systems.
 Conclusion: We revealed that the text mining and similar document search system were useful for developing a clinical practice guideline based on the evidence-based medicine. We could achieve high precision and recall rate through these systems. Above all, searches on the exercise that has a limited condition had an excellent result.