Students feedback analysis model using deep learning-based method and linguistic knowledge for intelligent educational systems

Abstract Student feedback analysis is time-consuming and laborious work if it handled manually. This study explores the use of a new deep learning-based method to design more accurate automated system for analysing students’ ( called DTLP: learning teaching process ). The DTLP employs convolutional neural networks (CNNs), bidirectional LSTM (BiLSTM), attention mechanism. To best our knowledge, using unified feature set, which representative word embedding, sentiment shifter rules, linguistic statistical has not been thoroughly studied with regard student feedback. Furthermore, uses multiple strategies overcome following drawbacks: contextual polarity; sentence types; words similar semantic context but opposite coverage limit an individual lexicon; sense variations . evaluate DTLP, we conducted experiment on large volume results showed i ) outperforms existing systems in field, ii that learns from this set can acquire significantly higher performance than one subset, iii ensemble statistical, linguistic, knowledge allows obtain significant performance, iv mechanism into CNN-BiLSTM improves DTLP. In addition, deployed looks potential causes behind