テキストマイニングによる薬学生実務実習レポートの分析

抄録

We have been conducting practical training for fourth-year students at Gifu Pharmaceutical University.At the end of the training,they write reports on their practical training in the hospital pharmacy and practical training in the community pharmacy.
We converted their sentence data into text format and performed Japanese-language morphologic analysis on it using language analysis software (KHCoder).We then worked out the frequency of keywords based on the word class information in the Japanese language morphologic analysis.We also classified the characteristics of keywords by assigning them to 5 categories which were“practice”,“patient”,“practical training”,“medical team”and“community healthcare”.
The results showed that the students’awareness of“patient” tended to be higher in the clinical pharmacy report.Further,the results of correspondence analysis showed that the characteristics of reports on training in the hospital pharmacy and reports on training in the community pharmacy were independent.We therefore considered text-mining to be a useful approach in the evaluation of practical training.