Hate speech detection in Twitter using hybrid embeddings and improved cuckoo search-based neural networks

2020 ◽  
Vol 13 (4) ◽  
pp. 485-525
Author(s):  
Femi Emmanuel Ayo ◽  
Olusegun Folorunso ◽  
Friday Thomas Ibharalu ◽  
Idowu Ademola Osinuga
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Feature Extraction ◽  
Hate Speech ◽  
Short Term Memory ◽  
Cuckoo Search ◽  
Research Attention ◽  
Content Type ◽  
Speech Detection ◽  
Sentence Level

PurposeHate speech is an expression of intense hatred. Twitter has become a popular analytical tool for the prediction and monitoring of abusive behaviors. Hate speech detection with social media data has witnessed special research attention in recent studies, hence, the need to design a generic metadata architecture and efficient feature extraction technique to enhance hate speech detection.Design/methodology/approachThis study proposes a hybrid embeddings enhanced with a topic inference method and an improved cuckoo search neural network for hate speech detection in Twitter data. The proposed method uses a hybrid embeddings technique that includes Term Frequency-Inverse Document Frequency (TF-IDF) for word-level feature extraction and Long Short Term Memory (LSTM) which is a variant of recurrent neural networks architecture for sentence-level feature extraction. The extracted features from the hybrid embeddings then serve as input into the improved cuckoo search neural network for the prediction of a tweet as hate speech, offensive language or neither.FindingsThe proposed method showed better results when tested on the collected Twitter datasets compared to other related methods. In order to validate the performances of the proposed method, t-test and post hoc multiple comparisons were used to compare the significance and means of the proposed method with other related methods for hate speech detection. Furthermore, Paired Sample t-Test was also conducted to validate the performances of the proposed method with other related methods.Research limitations/implicationsFinally, the evaluation results showed that the proposed method outperforms other related methods with mean F1-score of 91.3.Originality/valueThe main novelty of this study is the use of an automatic topic spotting measure based on naïve Bayes model to improve features representation.

Detecting hate speech against politicians in Arabic community on social media

2020 ◽  
Vol 16 (3) ◽  
pp. 295-313
Author(s):  
Imane Guellil ◽  
Ahsan Adeel ◽  
Faical Azouaou ◽  
Sara Chennoufi ◽  
Hanene Maafi ◽  
...  
Keyword(s):  
Social Media ◽  
Deep Learning ◽  
Hate Speech ◽  
Short Term Memory ◽  
Arabic Language ◽  
Short Term ◽  
Speech Corpus ◽  
Term Memory ◽  
Content Type ◽  
Speech Detection

Purpose This paper aims to propose an approach for hate speech detection against politicians in Arabic community on social media (e.g. Youtube). In the literature, similar works have been presented for other languages such as English. However, to the best of the authors’ knowledge, not much work has been conducted in the Arabic language. Design/methodology/approach This approach uses both classical algorithms of classification and deep learning algorithms. For the classical algorithms, the authors use Gaussian NB (GNB), Logistic Regression (LR), Random Forest (RF), SGD Classifier (SGD) and Linear SVC (LSVC). For the deep learning classification, four different algorithms (convolutional neural network (CNN), multilayer perceptron (MLP), long- or short-term memory (LSTM) and bi-directional long- or short-term memory (Bi-LSTM) are applied. For extracting features, the authors use both Word2vec and FastText with their two implementations, namely, Skip Gram (SG) and Continuous Bag of Word (CBOW). Findings Simulation results demonstrate the best performance of LSVC, BiLSTM and MLP achieving an accuracy up to 91%, when it is associated to SG model. The results are also shown that the classification that has been done on balanced corpus are more accurate than those done on unbalanced corpus. Originality/value The principal originality of this paper is to construct a new hate speech corpus (Arabic_fr_en) which was annotated by three different annotators. This corpus contains the three languages used by Arabic people being Arabic, French and English. For Arabic, the corpus contains both script Arabic and Arabizi (i.e. Arabic words written with Latin letters). Another originality is to rely on both shallow and deep leaning classification by using different model for extraction features such as Word2vec and FastText with their two implementation SG and CBOW.

scholarly journals Automated Amharic Hate Speech Posts and Comments Detection Model Using Recurrent Neural Network

2020 ◽  
Author(s):  
Surafel Getachew Tesfaye ◽  
Kula Kakeba
Keyword(s):  
Neural Network ◽  
Recurrent Neural Network ◽  
Performance Test ◽  
Hate Speech ◽  
Short Term Memory ◽  
Detection Methods ◽  
Batch Size ◽  
Neural Network Models ◽  
Data Set ◽  
Speech Detection

Abstract During the last few years, social activities over the internet especially on social media platforms increased drastically, but unfortunately, social networks have also become the place for hate speech proliferation by which most people’s social lives are disturbed because of hate speech posts and conflicts triggered by those posts. Studies confirm that online hate speech has different offline consequences. Even though there are a lot of researches on automated hate speech detection most of them are for other language and there is a scarcity of labeled data to apply automated analysis and detection methods on Amharic dataset. Therefore the research on automatic detection of hate speech posts attracted our attention. As a solution to those problems, this research aimed to prepare a labeled huge Amharic dataset by collecting posts and comments from selected Facebook pages of activists that participated actively. Those Facebook data sets are labeled manually as hate and free based on the guidelines given from researcher and pre-processed by applying data cleaning and normalization techniques. In this research the recurrent neural network models for automated hate speech posts detection from Amharic posts on Facebook is developed by using Long Short Term Memory (LSTM) and Gated Recurrent Unit (GRU) with word n-grams for feature extraction and word2vec to represent each unique word by vector representation. The experiment conducted on those two models by using 80% of the data set for training and 10% for validation to train the model and to select the best hyper-parameters combination for automated hate speech posts detection. The remaining 10% of the dataset used for testing the model after training. As a result LSTM based RNN of Batch size 128, and learning rate 0.001 with RMSProp optimizer and 0.5 dropout achieves an accuracy of 97.9% to detect posts as hate speech or free by training with 100 epochs. Which is assured by testing the models using models performance test and inference on user-generated data.

scholarly journals Classification of Skin Disease Using Deep Learning Neural Networks with MobileNet V2 and LSTM

Sensors ◽  
2021 ◽  
Vol 21 (8) ◽  
pp. 2852
Author(s):  
Parvathaneni Naga Srinivasu ◽  
Jalluri Gnana SivaSai ◽  
Muhammad Fazal Ijaz ◽  
Akash Kumar Bhoi ◽  
Wonjoon Kim ◽  
...  
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Deep Learning ◽  
Convolutional Neural Network ◽  
Skin Disease ◽  
Network Architecture ◽  
Large Scale ◽  
Short Term Memory ◽  
Convolutional Networks ◽  
Occurrence Matrix

Deep learning models are efficient in learning the features that assist in understanding complex patterns precisely. This study proposed a computerized process of classifying skin disease through deep learning based MobileNet V2 and Long Short Term Memory (LSTM). The MobileNet V2 model proved to be efficient with a better accuracy that can work on lightweight computational devices. The proposed model is efficient in maintaining stateful information for precise predictions. A grey-level co-occurrence matrix is used for assessing the progress of diseased growth. The performance has been compared against other state-of-the-art models such as Fine-Tuned Neural Networks (FTNN), Convolutional Neural Network (CNN), Very Deep Convolutional Networks for Large-Scale Image Recognition developed by Visual Geometry Group (VGG), and convolutional neural network architecture that expanded with few changes. The HAM10000 dataset is used and the proposed method has outperformed other methods with more than 85% accuracy. Its robustness in recognizing the affected region much faster with almost 2× lesser computations than the conventional MobileNet model results in minimal computational efforts. Furthermore, a mobile application is designed for instant and proper action. It helps the patient and dermatologists identify the type of disease from the affected region’s image at the initial stage of the skin disease. These findings suggest that the proposed system can help general practitioners efficiently and effectively diagnose skin conditions, thereby reducing further complications and morbidity.

scholarly journals Learners Demographics Classification on MOOCs During the COVID-19: Author Profiling via Deep Learning Based on Semantic and Syntactic Representations

2021 ◽  
Vol 6 ◽  
Author(s):  
Tahani Aljohani ◽  
Alexandra I. Cristea
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Deep Learning ◽  
Prediction Models ◽  
Short Term Memory ◽  
Methodological Approach ◽  
High Accuracy ◽  
Directional Model ◽  
Textual Representations ◽  
The One

Massive Open Online Courses (MOOCs) have become universal learning resources, and the COVID-19 pandemic is rendering these platforms even more necessary. In this paper, we seek to improve Learner Profiling (LP), i.e. estimating the demographic characteristics of learners in MOOC platforms. We have focused on examining models which show promise elsewhere, but were never examined in the LP area (deep learning models) based on effective textual representations. As LP characteristics, we predict here the employment status of learners. We compare sequential and parallel ensemble deep learning architectures based on Convolutional Neural Networks and Recurrent Neural Networks, obtaining an average high accuracy of 96.3% for our best method. Next, we predict the gender of learners based on syntactic knowledge from the text. We compare different tree-structured Long-Short-Term Memory models (as state-of-the-art candidates) and provide our novel version of a Bi-directional composition function for existing architectures. In addition, we evaluate 18 different combinations of word-level encoding and sentence-level encoding functions. Based on these results, we show that our Bi-directional model outperforms all other models and the highest accuracy result among our models is the one based on the combination of FeedForward Neural Network and the Stack-augmented Parser-Interpreter Neural Network (82.60% prediction accuracy). We argue that our prediction models recommended for both demographics characteristics examined in this study can achieve high accuracy. This is additionally also the first time a sound methodological approach toward improving accuracy for learner demographics classification on MOOCs was proposed.

scholarly journals Part-of-Speech Tagging via Deep Neural Networks for Northern-Ethiopic Languages

2020 ◽  
Vol 49 (4) ◽  
pp. 482-494
Author(s):  
Jurgita Kapočiūtė-Dzikienė ◽  
Senait Gebremichael Tesfagergish
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Language Processing ◽  
Deep Neural Networks ◽  
Short Term Memory ◽  
Parameter Tuning ◽  
Feed Forward Neural Network ◽  
Pos Tagging ◽  
Part Of Speech ◽  
Pos Tagger

Deep Neural Networks (DNNs) have proven to be especially successful in the area of Natural Language Processing (NLP) and Part-Of-Speech (POS) tagging—which is the process of mapping words to their corresponding POS labels depending on the context. Despite recent development of language technologies, low-resourced languages (such as an East African Tigrinya language), have received too little attention. We investigate the effectiveness of Deep Learning (DL) solutions for the low-resourced Tigrinya language of the Northern-Ethiopic branch. We have selected Tigrinya as the testbed example and have tested state-of-the-art DL approaches seeking to build the most accurate POS tagger. We have evaluated DNN classifiers (Feed Forward Neural Network – FFNN, Long Short-Term Memory method – LSTM, Bidirectional LSTM, and Convolutional Neural Network – CNN) on a top of neural word2vec word embeddings with a small training corpus known as Nagaoka Tigrinya Corpus. To determine the best DNN classifier type, its architecture and hyper-parameter set both manual and automatic hyper-parameter tuning has been performed. BiLSTM method was proved to be the most suitable for our solving task: it achieved the highest accuracy equal to 92% that is 65% above the random baseline.

scholarly journals Music generation and human voice conversion based on LSTM

2021 ◽  
Vol 336 ◽  
pp. 06015
Author(s):  
Guangwei Li ◽  
Shuxue Ding ◽  
Yujie Li ◽  
Kangkang Zhang
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Short Term Memory ◽  
Human Life ◽  
Piano Music ◽  
Training Model ◽  
Voice Conversion ◽  
Neural Network Models ◽  
Human Voice ◽  
Music Generation

Music is closely related to human life and is an important way for people to express their feelings in life. Deep neural networks have played a significant role in the field of music processing. There are many different neural network models to implement deep learning for audio processing. For general neural networks, there are problems such as complex operation and slow computing speed. In this paper, we introduce Long Short-Term Memory (LSTM), which is a circulating neural network, to realize end-to-end training. The network structure is simple and can generate better audio sequences after the training model. After music generation, human voice conversion is important for music understanding and inserting lyrics to pure music. We propose the audio segmentation technology for segmenting the fixed length of the human voice. Different notes are classified through piano music without considering the scale and are correlated with the different human voices we get. Finally, through the transformation, we can express the generated piano music through the output of the human voice. Experimental results demonstrate that the proposed scheme can successfully obtain a human voice from pure piano Music generated by LSTM.

scholarly journals Pattern Recognition by Hierarchical Feature Extraction

2003 ◽  
Vol 15 (3) ◽  
pp. 278-285
Author(s):  
Daigo Misaki ◽  
◽  
Shigeru Aomura ◽  
Noriyuki Aoyama
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Pattern Recognition ◽  
Feature Extraction ◽  
Input Data ◽  
Local Features ◽  
Detailed Classification

We discuss effective pattern recognition for contour images by hierarchical feature extraction. When pattern recognition is done for an unlimited object, it is effective to see the object in a perspective manner at the beginning and next to see in detail. General features are used for rough classification and local features are used for a more detailed classification. D-P matching is applied for classification of a typical contour image of individual class, which contains selected points called ""landmark""s, and rough classification is done. Features between these landmarks are analyzed and used as input data of neural networks for more detailed classification. We apply this to an illustrated referenced book of insects in which much information is classified hierarchically to verify the proposed method. By introducing landmarks, a neural network can be used effectively for pattern recognition of contour images.

scholarly journals Classification and direction discrimination of faults in transmission lines using 1D convolutional neural networks

2021 ◽  
Vol 12 (3) ◽  
pp. 1928
Author(s):  
Ahmed Thamer Radhi ◽  
Wael Hussein Zayer ◽  
Adel Manaa Dakhil
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Feature Extraction ◽  
Transmission Line ◽  
Convolutional Neural Networks ◽  
Transmission Lines ◽  
Fault Classification ◽  
Line System ◽  
Direction Discrimination ◽  
Conventional Methods

<span lang="EN-US">This paper presents a fast and accurate fault detection, classification and direction discrimination algorithm of transmission lines using one-dimensional convolutional neural networks (1D-CNNs) that have ingrained adaptive model to avoid the feature extraction difficulties and fault classification into one learning algorithm. A proposed algorithm is directly usable with raw data and this deletes the need of a discrete feature extraction method resulting in more effective protective system. The proposed approach based on the three-phase voltages and currents signals of one end at the relay location in the transmission line system are taken as input to the proposed 1D-CNN algorithm. A 132kV power transmission line is simulated by Matlab simulink to prepare the training and testing data for the proposed 1D- CNN algorithm. The testing accuracy of the proposed algorithm is compared with other two conventional methods which are neural network and fuzzy neural network. The results of test explain that the new proposed detection system is efficient and fast for classifying and direction discrimination of fault in transmission line with high accuracy as compared with other conventional methods under various conditions of faults.</span>

scholarly journals The Role and Utilization of CNN in Automatic Logo Based Document Image Retrieval Methods

2018 ◽  
Vol 7 (3.1) ◽  
pp. 13
Author(s):  
Raveendra K ◽  
R Vinoth Kanna
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Feature Extraction ◽  
Image Retrieval ◽  
Convolutional Neural Network ◽  
Convolutional Neural Networks ◽  
Document Image ◽  
Retrieval Process ◽  
Pictorial Representations

Automatic logo based document image retrieval process is an essential and mostly used method in the feature extraction applications. In this paper the architecture of Convolutional Neural Network (CNN) was elaborately explained with pictorial representations in order to understand the complex Convolutional Neural Networks process in a simplified way. The main objective of this paper is to effectively utilize the CNN in the process of automatic logo based document image retrieval methods.  

Forecasting the momentum using customised loss function for financial series

2021 ◽  
Vol 14 (4) ◽  
pp. 702-713
Author(s):  
N. Prabakaran ◽  
Rajasekaran Palaniappan ◽  
R. Kannadasan ◽  
Satya Vinay Dudi ◽  
V. Sasidhar
Keyword(s):  
Neural Network ◽  
Time Series ◽  
Loss Function ◽  
Short Term Memory ◽  
Learning Algorithm ◽  
Financial Time Series ◽  
Storage Unit ◽  
Content Type ◽  
Financial Time ◽  
Unit Structure

PurposeWe propose a Machine Learning (ML) approach that will be trained from the available financial data and is able to gain the trends over the data and then uses the acquired knowledge for a more accurate forecasting of financial series. This work will provide a more precise results when weighed up to aged financial series forecasting algorithms. The LSTM Classic will be used to forecast the momentum of the Financial Series Index and also applied to its commodities. The network will be trained and evaluated for accuracy with various sizes of data sets, i.e. weekly historical data of MCX, GOLD, COPPER and the results will be calculated.Design/methodology/approachDesirable LSTM model for script price forecasting from the perspective of minimizing MSE. The approach which we have followed is shown below. (1) Acquire the Dataset. (2) Define your training and testing columns in the dataset. (3) Transform the input value using scalar. (4) Define the custom loss function. (5) Build and Compile the model. (6) Visualise the improvements in results.FindingsFinancial series is one of the very aged techniques where a commerce person would commerce financial scripts, make business and earn some wealth from these companies that vend a part of their business on trading manifesto. Forecasting financial script prices is complex tasks that consider extensive human–computer interaction. Due to the correlated nature of financial series prices, conventional batch processing methods like an artificial neural network, convolutional neural network, cannot be utilised efficiently for financial market analysis. We propose an online learning algorithm that utilises an upgraded of recurrent neural networks called long short-term memory Classic (LSTM). The LSTM Classic is quite different from normal LSTM as it has customised loss function in it. This LSTM Classic avoids long-term dependence on its metrics issues because of its unique internal storage unit structure, and it helps forecast financial time series. Financial Series Index is the combination of various commodities (time series). This makes Financial Index more reliable than the financial time series as it does not show a drastic change in its value even some of its commodities are affected. This work will provide a more precise results when weighed up to aged financial series forecasting algorithms.Originality/valueWe had built the customised loss function model by using LSTM scheme and have experimented on MCX index and as well as on its commodities and improvements in results are calculated for every epoch that we run for the whole rows present in the dataset. For every epoch we can visualise the improvements in loss. One more improvement that can be done to our model that the relationship between price difference and directional loss is specific to other financial scripts. Deep evaluations can be done to identify the best combination of these for a particular stock to obtain better results.

Sign in / Sign up

Export Citation Format

Share Document