利用自然语言处理技术提取致病因素信息研究

Research on Extraction Epidemiologic Exposures Based on Natural Language Processing Techniques

基于自然语言处理技术,设计出能够自动提取流行病学致病因素的系统。经过评估发现该系统由一个自然语言处理子系统和一个基于逻辑回归算法或支持向量机算法的信息提取子系统构成。可以高效地从生物医学文献中自动获取致病因素信息。

Based on natural language processing techniques, a system is designed automatically extract epidemiologic exposures. E- valuation of the system shows that it consists of a natural language processing subsystem and a logistic regression based or a support vector machine based information extraction subsystem. The system can efficiently help epidemiologic researchers get exposures automatically from biomedical literature.