Sentence-Level Classification Using Parallel Fuzzy Deep Learning Classifier

Abstract

At present, with the growing number of Web 2.0 platforms such as Instagram, Facebook, and Twitter, users honestly communicate their opinions and ideas about events, services, and products. Owing to this rise in the number of social platforms and their extensive use by people, enormous amounts of data are produced hourly. However, sentiment analysis or opinion mining is considered as a useful tool that aims to extract the emotion and attitude from the user-posted data on social media platforms by using different computational methods to linguistic terms and various Natural Language Processing (NLP). Therefore, enhancing text sentiment classification accuracy has become feasible, and an interesting research area for many community researchers. In this study, a new Fuzzy Deep Learning Classifier (FDLC) is suggested for improving the performance of data-sentiment classification. Our proposed FDLC integrates Convolutional Neural Network (CNN) to build an effective automatic process for extracting the features from collected unstructured data and Feedforward Neural Network (FFNN) to compute both positive and negative sentimental scores. Then, we used the Mamdani Fuzzy System (MFS) as a fuzzy classifier to classify the outcomes of the two used deep (CNN+FFNN) learning models in three classes, which are: Neutral, Negative, and Positive. Also, to prevent the long execution time taking by our hybrid proposed FDLC, we have implemented our proposal under the Hadoop cluster. An experimental comparative study between our FDLC and some other suggestions from the literature is performed to demonstrate our offered classifier's effectiveness. The empirical result proved that our FDLC performs better than other classifiers in terms of true positive rate, true negative rate, false positive rate, false negative rate, error rate, precision, classification rate, kappa statistic, F1-score and time consumption, complexity, convergence, and stability.

A MapReduce Opinion Mining for COVID-19-Related Tweets Classification Using Enhanced ID3 Decision Tree Classifier

“…In [67], the authors proposed a new hybrid approach called IGCN that integrates a bidirectional gating mechanism and CNN for prophesying the opinion of a target. Es-Sabery et al [68] proposed a new fuzzy deep learning classifier for performing opinion mining. This hybrid approach combines the feedforward and convolutional neural network (FFNN + CNN) and Mamdani fuzzy system (MFS).…”

“…To assess our text classification method, we principally compute ten assessment metrics: True Positive Rate (TPR), True Negative Rate (TNR), Kappa Statistic (KS), False Positive Rate (FPR), Precision (PR), False Negative Rate (FNR), Classification Rate or Accuracy (AC), Error Rate (ER), Time Consumption (TC), and F1-score (FS) [103]. These evaluation metrics are computed as described in Table 7 and based on the confusion matrix for binary classification [68] as given in Fig. 9.…”

Sentiment Analysis of Covid19 Tweets Using A MapReduce Fuzzified Hybrid Classifier Based On C4.5 Decision Tree and Convolutional Neural Network

“…In this work, we apply the fuzzification function to turn out the neuron values of denser layer to a set of fuzzy values by measuring the membership degree of each neuron value employing Gaussian membership function. We have chosen the Gaussian membership function instead of triangular of trapezoidal membership functions because of the experimental result presented in the paper [3], which proved that the Gaussian membership function achieves a good accuracy equal to 94.87% compared to trapezoidal membership function that reaches an accuracy equal to 91.21% and triangular membership function that gives an accuracy equal to 90.14%. The Gaussian function is defined by two variables r is the central value, and d > 0 indicates the standard deviation and the membership degree of variable z is computed employing the next equation ( 5).…”

“…Although applying the most efficient machine learning and deep learning approaches, NLP's inherent vagueness requires more solutions. Numerous works from the literature [3], [4][5] prove that fuzzy logic theories are the appropriate techniques to handle ambiguous, uncertain and imprecise information.…”

Remote Sensing Image Fusion Based on Nonnegative Dictionary Learning

“…Most of the above algorithms need to use the similar high-resolution multispectral image samples to train the dictionary, but the similar high-resolution multispectral image is often difficult to obtain, which limits the application of the algorithm in practice. In addition, with the rise of deep learning technology [20][21][22], this technology is also used in remote sensing image fusion [23][24][25], but this kind of algorithm needs a large number of samples for training.…”