摘要：
Sentiment recognition of online course reviews is valuable to understand emotions and feelings of learners. Nowadays, an increasing number of course reviews are being generated with the emergence of Massive Open Online Courses (MOOCs), which offers teachers a chance to analyze the opinions of learners and improve teaching strategies. However, the unstructured data contain large amounts of redundant features, which will significantly impact the performance of machine learning. To select effective emotional features, we adopt a multi-swarm particle swarm optimization (MSPSO) method, which generates multi diverse particle swarms on several cross training subsets. These swarms are utilized to find the best features by the F-Measure fitness function. The experimental results on the real-life dataset show that MSPSO can effectively reduce redundancy of text features and capture discriminative features. Compared with conventional feature selection methods, MSPSO can gain the better performance when selecting the same dimensions. Besides, the result of a user survey indicates that 72.19% of subjects approve of the usability of the recognition results and effectiveness of the feature selection. MSPSO is effective to pick discriminative features in course reviews.Discriminability of features is used to iteratively optimize particle swarms.MSPSO exploits the differences among sample subsets to form diverse swarms. Sentiment recognition of online course reviews is valuable to understand emotions and feelings of learners. Nowadays, an increasing number of course reviews are being generated with the emergence of Massive Open Online Courses (MOOCs), which offers teachers a chance to analyze the opinions of learners and improve teaching strategies. However, the unstructured data contain large amounts of redundant features, which will significantly impact the performance of machine learning. To select effective emotional features, we adopt a multi-swarm particle swarm optimization (MSPSO) method, which generates multi diverse particle swarms on several cross training subsets. These swarms are utilized to find the best features by the F-Measure fitness function. The experimental results on the real-life dataset show that MSPSO can effectively reduce redundancy of text features and capture discriminative features. Compared with conventional feature selection methods, MSPSO can gain the better performance when selecting the same dimensions. Besides, the result of a user survey indicates that 72.19% of subjects approve of the usability of the recognition results and effectiveness of the feature selection. MSPSO is effective to pick discriminative features in course reviews.Discriminability of features is used to iteratively optimize particle swarms.MSPSO exploits the differences among sample subsets to form diverse swarms.
作者：
Liu, Zhi*;Zhang, Wenjing;Sun, Jianwen（孙建文）;Cheng, Hercy N. H.;Peng, Xian;...
作者机构：
[Sun, Jianwen; Liu, Zhi; Cheng, Hercy N. H.; Peng, Xian; Liu, Sanya; Zhang, Wenjing] Cent China Normal Univ, Natl Engn Res Ctr E Learning, Wuhan, Peoples R China.
会议名称：
5th International Conference on Educational Innovation through Technology (EITT)
会议时间：
SEP 22-24, 2016
会议地点：
Tainan, TAIWAN
会议主办单位：
[Liu, Zhi;Zhang, Wenjing;Sun, Jianwen;Cheng, Hercy N. H.;Peng, Xian;Liu, Sanya] Cent China Normal Univ, Natl Engn Res Ctr E Learning, Wuhan, Peoples R China.
关键词：
Massive Open Online Course (MOOC);course comments;emotion recognition;topic mining;learning analytics
摘要：
Massive Open Online Course (MOOC) has been drawn much attention from learners and teachers through the world. MOOC offers a variety of interactive ways, in which the course comment panel is used for express students' opinions and feelings. These comments generally contain some learning problems, attitudes towards the course or the platform support, etc. The feedback information is beneficial for the exchange of ideas among teachers, learners and educational administrators. However, it is quite time-consuming to analyze these important opinions entirely by artificial reading. It is imperative that the MOOC needs the machine learning methods to detect the emotions and topics in text data. In this paper, we propose an application framework and design scheme of intelligent system for the emotion recognition and topic mining, aiming at conducting the intelligent and personalized learning analytics on MOOC. The purposes of the intelligent comment mining system include (1) predicting popularity level of each course; (2) obtaining emotion-topic feedbacks about content of courses for teachers to analyze and improve their teaching strategies; (3) obtaining emotion-topic feedbacks about platform support for administrators to improve user experiences in platform.
期刊：
International Journal of Computers and Applications,2015年37(3-4):94-101 ISSN：1206-212X
通讯作者：
Liu, Sanya(lsy5918@gmail.com)
作者机构：
[Jianwen Sun; Zhi Liu; Sanya Liu; Lin Liu; Meng Wang; Xian Peng] National Engineering Research Center for E-Learning, Central China Normal University, Wuhan, P.R. China
通讯机构：
[Sanya Liu] N;National Engineering Research Center for E-Learning, Central China Normal University, Wuhan, P.R. China
摘要：
A blind deconvolution algorithm with modified Tikhonov regularization is introduced. To improve the spectral resolution, spectral structure information is incorporated into regularization by using the adaptive term to distinguish the spectral structure from other regions. The proposed algorithm can effectively suppress Poisson noise as well as preserve the spectral structure and detailed information. Moreover, it becomes more robust with the change of the regularization parameter. Comparative results on simulated and real degraded Raman spectra are reported. The recovered Raman spectra can easily extract the spectral features and interpret the unknown chemical mixture.
作者：
Jianwen Sun（孙建文）;Zongkai Yang;Sanya Liu;Pei Wang 0010
期刊：
Journal of Networks,2012年7(2):259-266 ISSN：1796-2056
通讯作者：
Liu, S.(lsy5918@gmail.com)
作者机构：
[Jianwen Sun; Zongkai Yang; Sanya Liu] National Engineering Research Center for E-learning, Central China Normal University, Wuhan, China;[Pei Wang 0010] School of Information Management, Wuhan University, Wuhan, China
作者：
Liu, Sanya;Liu, Zhi;Sun, Jianwen（孙建文）;Liu, Lin
期刊：
International Journal of Digital Content Technology and its Applications,2011年5(3):126-135 ISSN：1975-9339
通讯作者：
Liu, S.(lsy5918@gmail.com)
作者机构：
[Sun, Jianwen; Liu, Sanya; Liu, Zhi; Liu, Lin] Engineering and Research Center for Information Technology on Education, Central China Normal University, National Engineering Research Center for E-Learning, Hubei Wuhan, China
摘要：
Synergetic neural network (SNN) associates synergetics with artificial neural network, it can rigorously deal with the behavior of network in the mathematical theory, and have the advantage of fast learning, short pattern recalling time and so on. In this paper, a pattern recognition method based on the self-adaptive attention parameters presented on the basis of analyzing the key technology of SNN, and the advanced algorithm will be employed in the online writeprint identification, the key point of this algorithm is that it can correct initial mis-identified patterns through measuring similarity between the prototype pattern and the testing pattern in the evolution of order parameters. Experimental results show that the advanced SNN has better performance and robustness than the SNN based on balanced attention parameters. Further, the network's self-learning ability and recognition performance is greatly improved by using advanced SNN.
