TY - GEN
T1 - ELMo-ACSA based Multi-aspect View Mining of Students' Reviews on Teaching
AU - Wang, Yuanyuan
AU - Yang, Zichen
AU - Tian, Feng
AU - Wu, Fan
AU - Chen, Yan
AU - Zheng, Qinghua
N1 - Publisher Copyright:
© 2019 IEEE.
PY - 2019/7
Y1 - 2019/7
N2 - student reviews on teaching encompasses multiple teaching assessment aspects and different emotional tendencies. In this paper, we propose a multi-aspects view mining method for student reviews on teaching, combined with the teaching assessment index system of Xi'an Jiaotong University. The proposed method aims to mine the detailed and rich emotional information of student reviews on teaching, find the effectiveness of teachers and classes, identify the problems encountered classroom accurately, and provide technical support for intelligent teaching feedback. The multi-aspect view mining is divided into two parts: teaching assessment aspect definition and sentiment classification aspect, which are respectively based on the multi-level evaluation index of Xi'an Jiaotong University and the combination of pre-training language model and gated convolutional networks. Experimental verification was carried out on the longer assessment reviews collected from the big data platform for real-time monitoring of teaching quality. The F value of the proposed method is superior to the comparison model, indicating the effectiveness of the multi-aspect view mining method.
AB - student reviews on teaching encompasses multiple teaching assessment aspects and different emotional tendencies. In this paper, we propose a multi-aspects view mining method for student reviews on teaching, combined with the teaching assessment index system of Xi'an Jiaotong University. The proposed method aims to mine the detailed and rich emotional information of student reviews on teaching, find the effectiveness of teachers and classes, identify the problems encountered classroom accurately, and provide technical support for intelligent teaching feedback. The multi-aspect view mining is divided into two parts: teaching assessment aspect definition and sentiment classification aspect, which are respectively based on the multi-level evaluation index of Xi'an Jiaotong University and the combination of pre-training language model and gated convolutional networks. Experimental verification was carried out on the longer assessment reviews collected from the big data platform for real-time monitoring of teaching quality. The F value of the proposed method is superior to the comparison model, indicating the effectiveness of the multi-aspect view mining method.
KW - ELMo
KW - Gated Convolutional Networks
KW - Multi-aspects Sentimation Classification
KW - Students' Reviews on Teaching
KW - View Mining
UR - https://www.scopus.com/pages/publications/85071916419
U2 - 10.1109/BigDIA.2019.8802779
DO - 10.1109/BigDIA.2019.8802779
M3 - 会议稿件
AN - SCOPUS:85071916419
T3 - Proceedings - 2019 5th International Conference on Big Data and Information Analytics, BigDIA 2019
SP - 78
EP - 83
BT - Proceedings - 2019 5th International Conference on Big Data and Information Analytics, BigDIA 2019
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 5th International Conference on Big Data and Information Analytics, BigDIA 2019
Y2 - 8 July 2019 through 10 July 2019
ER -