著者
廣川 佐千男 中藤 哲也 殷 成久 鈴木 孝彦 橋本 喜代太
雑誌
人工知能学会全国大会
巻号頁・発行日
vol.26, no.1, pp.1-2, 2012-06-12

Since evaluation of food is subjective and the feeling is expressed in Japanese, onomatopoeia is used in many cases. In cooking and sweets, the onomatopoeia used is different. In order to grasp the meaning of onomatopoeia, we need to read and compare the situation where each onomatopoeia is used. In this report, the "Double Rank" method which the first author proposed is applied to Blog documents to compare and to analyze Blogs.set of all words gains the best performance if 90% of the data are used as a training data. However, the set of a small number of words with positive scores outperforms other feature sets, if the training data is only 10%. In such a realistic situation, the feature words are effective in improving discernment.