Augmented language model with deep learning adaptation on sentiment analysis for E-learning recommendation

Sentiment Analysis is considered as an important research field in text mining, and is significant in recommendation systems and e-learning environments. This research proposes a new methodology of e-learning hybrid Recommendation System Based on Sentiment Analysis (RSBSA) by leveraging tailored Natural Language Processing (NLP) and Convolutional Neural Network (CNN) techniques, to recommend appropriate e-learning materials based on learner’s preferences. Integration is done on fine-grained sentiment analysis models, to classify text reviews of e-content posted on e-learning platform. Two enhanced language models based on ‘Continuous Bag of Word’ and ‘Skip-Gram’ are introduced. Moreover, three resilient language models based on the hybrid language techniques are developed to produce a superior vocabulary representation. These models were trained using various CNN models to predict ratings of resources from online reviews provided by learners. To accomplish this, a customizable dataset ‘ABHR-1′ is used, which is derived from e-content' reviews with corresponding ratings labeled [1–5]. The proposed models are evaluated and tested using ABHR-1 and two public datasets. According to the simulation results, Multiplication-Several-Channels-CNN model outperformed other models with an accuracy of 90.37 % for fine-grained sentiment classification on 5 discrete classes and the empirical results are compared.     

Investigation of psycho-physiological interactions between patient and therapist during a psychodynamic therapy and their relation to speech using in terms of entropy analysis using a neural network approach

In the present contribution we investigate in an exemplary single-case study the behavior of psycho-physiological variables in psychotherapy sessions. The values are measured continously during a single session at the same time for both patient and therapist. The analysis of the data is done using an artificial neural network approach for non-linear principal component analysis and faithful data representation/visualization and compression required for subsequent process analysis. The used network (growing self-organizing map, GSOM) thereby uses a kernel smoothing for improved data density estimation. In this way, we are able to generate an entropy model of psycho-physiological variability detecting emotionally instable phases during the therapy process. We relate our finding to results obtained by speech analysis of the therapy sessions according to the cycle model invented by Mergenthaler. Thus, we get preliminary suggestions how psycho-physiological reactions are related to the therapeutic process.