F-019
Interview Mining using Machine Learning and Visualization

◎鄒　楽晗・金川久美子・戸崎祐輔・中藤哲也・廣川佐千男（九大）

We are analyzing the interview records of 36 different practicing librarians, archivists and museum curators across the world who specialize in East Asian collections, based on text mining and visualization. The present paper reports a keyword map system and case studies on the characteristic words for required ability of curators. The keyword map is obtained by inputting q query in the system. Co-occurrence relation of words is used to analyze the characteristics of words in questions and answers. The findings can accurately identify the answers that the interviewer wants according to the keywords in the question.