A teaching evaluation method based on sentiment classification. (2016)

• Record Type: Journal Article
• Authors: Zhao, Hua; Ji, Xiaowen; Zeng, Qingtian; Jiang, Shan
• Is Part Of: International journal of computing science and mathematics. Volume 7:Number 1(2016)
• Page Start: 54
• Page End: 62
• Publication Date: 2016
• Subjects: sentiment classification -- teaching evaluation -- evaluation word recognition -- sentiment dictionary -- sentiment analysis -- new words; Mathematics -- Periodicals; Computer science -- Periodicals; Mathematics -- Data processing -- Periodicals; 510.285
• Languages: English
• ISSNs: 1752-5055

Abstract

Teaching evaluation is an important part of the teaching process, and is an effective measure to improve teaching method. In order to automatically analyse the massive teaching evaluation texts on internet, this paper proposes to apply the sentiment classification technology to the sentiment analysis of teaching evaluation texts, and introduces a teaching evaluation analysis method based on the sentiment dictionary. In order to deal with the problem of the frequent appearance of new words in these texts, propose a method to recognise the new words automatically, and then use these recognised words to expand the sentiment dictionary. Experimental results show that the sentiment classification based on the expanded sentiment dictionary increases the recall rate of the system successfully, which brings the improvement of the overall performance of the teaching evaluation analysis system.