Sentiment Analysis for Distance Education Course Materials: A Machine Learning Approach

Abstract

Nowadays many companies and institutions are interested in learning what do people think and want. Many studies are conducted to answer these questions. That's why, emotions of people are significant in terms of instructional design. However, processing and analysis of many people's ideas and emotions is a challenging task. That is where the 'sentiment analysis' through machine learning techniques steps in. Recently a fast digitalization process is witnessed. Anadolu university, that serves 1 million distant students, is trying to find its place in this digital era. A learning management system (LMS) that distant students of the Open Education Faculty (Açıköğretim Fakültesi) is developed at the Anadolu University. Interaction with students is the clear advantage of LMS's when compared to the hard copy materials. Book, audio book (mp3), video and interactive tests are examples of these materials. 6059 feedbacks for those online materials was scaled using the triple Likert method and using machine learning techniques sentiment analysis was performed in this study. 0.775 correctness ratio was achieved via the Logistic regression algorithm. The research concludes that machine learning techniques can be used to better understand learners and how they feel.

Prediction of Survival and Recurrence Patterns by Machine Learning in Gastric Cancer Cases Undergoing Radiation Therapy and Chemotherapy

“…Logistic regression is produced using the natural logarithm of the probabilities of the target variable. 10 Artificial neural networks-MLP are developed based on the biological neural networks of the human brain and an information processing system designed to perform the functions of these networks. Artificial neural networks collect their knowledge by detecting patterns and relationships in the data and learn by experience.…”

A U‐Net Approach to Apical Lesion Segmentation on Panoramic Radiographs

“…This matrix is a meaningful table that summarizes the predicted and actual situations. The performance of model is frequently assessed using the data in the confusion matrix [27]. The metrics used to evaluate the success of the model were as follows: (1) True Positive (TP): apical lesion was segmented, correctly 3-5).…”

“…This matrix is a meaningful table that summarizes the predicted and actual situations. The performance of model is frequently assessed using the data in the confusion matrix [ 27 ]. The metrics used to evaluate the success of the model were as follows: True Positive (TP): apical lesion was segmented, correctly False Positive (FP): apical lesions were not detected False Negative (FN): without apical lesions, lesions were nevertheless segmented TP, FP, and FN were determined; then, the following metrics were computed: Sensitivity (recall): TP/(TP + FN) Precision: TP/(TP + FP) F1 score: 2TP/(2TP + FP + FN) …”

Hybrid neural network model based on multi-head attention for English text emotion analysis

“…Although these neural network models have achieved great success, they are difficult to extract multi-level and more comprehensive emotion features of text, and rely heavily on text information and emotion resources. Language knowledge [12] (emotion dictionary, negative words, adverbs of degree) needs to be integrated into the model to achieve the best potential in terms of prediction accuracy [13]. With the advent of capsules [14], Wang [15] first attempted to conduct emotion analysis through capsules, which did not require any assistance of language knowledge and had a higher classification accuracy compared to the baseline model.…”

“…The global average pooling method is used to integrate the local features and global semantic features of text to obtain the text instance feature representation s V , which enhances the feature expression ability of the model.During the experiment, the number of convolution kernel B in CNN is set to the same value as the output vector dimension 2d of the Bi-GRU network. The generated feature vectors by the two networks are combined and spliced as shown in equation(13). the output vector of the bidirectional GRU.…”