Identifikasi Hubungan Sebab-Akibat pada Artikel Kesehatan menggunakan Anotasi Elemen Medis dan Paragraf

Abstract

This paper studies natural language processing on medical articles in Indonesian that aims to identify causal relationship and used as public health surveillance information monitoring system. This paper proposes selection-feature conformity, phrase annotation, paragraph annotation, and medical element annotation. System performance evaluation is carried out using intrinsic aprroach which compares supervised classification methods, i.e. naive bayes method and HMM. Results obtained for recall, precission, and f-measure are 0.905, 0.924, 0.910 and 0.706, 0.750, 0.720, respectively

Keywords

• Sebab-akibat;artikel kesehatan;medical named entities;seleksi fitur;anotasi elemen medis;anotasi paragraf