Related Experiment Videos
Locative inferences in medical texts.
P S Mayer1, G H Bailey, R J Mayer
1Department of Industrial Engineering, Texas A&M University, College Station 77843.
Journal of Medical Systems
|June 1, 1987
Summary
This study introduces a novel approach to understanding clinical records for epidemiological research. It focuses on locative inferences, integrating temporal, locative, and conceptual data for improved text analysis.
Area of Science:
- Medical Informatics
- Computational Linguistics
- Epidemiology
Background:
- Epidemiological studies require large-scale clinical record analysis.
- Current clinical record data extraction faces challenges with temporal, locative, and conceptual information.
- Existing research has focused on temporal and conceptual inferences, neglecting locative aspects.
Purpose of the Study:
- To examine locative inferences in clinical record understanding.
- To integrate temporal, locative, and conceptual information for improved data extraction.
- To present an application demonstrating a novel parsing strategy for clinical records.
Main Methods:
- Developed a knowledge-based parsing strategy.
- Utilized a minimal lexicon for text analysis.
- Focused on extracting locative information alongside temporal and conceptual data.
Main Results:
- Demonstrated the feasibility of extracting locative inferences from clinical records.
- Showcased the integration of temporal, locative, and conceptual information processing.
- Presented a functional application for enhanced clinical record understanding.
Conclusions:
- Locative inference is crucial for comprehensive clinical record analysis.
- The proposed knowledge-based parsing strategy effectively integrates diverse information types.
- This research advances natural language processing applications in medical informatics and epidemiology.