Abstract
Currently, nursing-care freestyle texts (nursing-care data) are collected from many hospitals in
Japan via the Internet and stored into the database. For improving nursing care quality, experts need to
read all freestyle texts carefully. However, it is a hard task for each expert to evaluate the data because
of huge number of nursing-care data in the database. In order to reduce workloads to evaluate nursingcare
data, we have proposed a support vector machine (SVM) based classification system. In this paper,
to improve the classification performance, we propose a genetic algorithm (GA) based feature selection
method for generating numerical data from collected nursing-care texts. First, we extract nouns and verbs
from nursing-care texts using the morphological analysis software "MeCab" and store the extracted terms
into a "term list". Some combinations of terms in the term list are selected by GA with two objectives; (1)
maximization of the number of correctly classified texts and (2) minimization of the number of selected
terms. And then, we classify the nursing-care numerical data using the SVM. From computer simulation
results, we show the effectiveness of our proposed method.
