Multilingual Hate Speech Detection: A Semi-Supervised Generative Adversarial Approach.

Social media platforms have surpassed cultural and linguistic boundaries, thus enabling online communication worldwide. However, the expanded use of various languages has intensified the challenge of online detection of hate speech content. Despite the release of multiple Natural Language Processing (NLP) solutions implementing cutting-edge machine learning techniques, the scarcity of data, especially labeled data, remains a considerable obstacle, which further requires the use of semisupervised approaches along with Generative Artificial Intelligence (Generative AI) techniques. This paper introduces an innovative approach, a multilingual semisupervised model combining Generative Adversarial Networks (GANs) and Pretrained Language Models (PLMs), more precisely mBERT and XLM-RoBERTa. Our approach proves its effectiveness in the detection of hate speech and offensive language in Indo-European languages (in English, German, and Hindi) when employing only 20% annotated data from the HASOC2019 dataset, thereby presenting significantly high performances in each of multilingual, zero-shot crosslingual, and monolingual training scenarios. Our study provides a robust mBERT-based semisupervised GAN model (SS-GAN-mBERT) that outperformed the XLM-RoBERTa-based model (SS-GAN-XLM) and reached an average F1 score boost of 9.23% and an accuracy increase of 5.75% over the baseline semisupervised mBERT model.

Unveiling Human Values: Analyzing Emotions behind Arguments.

Detecting the underlying human values within arguments is essential across various domains, ranging from social sciences to recent computational approaches. Identifying these values remains a significant challenge due to their vast numbers and implicit usage in discourse. This study explores the potential of emotion analysis as a key feature in improving the detection of human values and information extraction from this field. It aims to gain insights into human behavior by applying intensive analyses of different levels of human values. Additionally, we conduct experiments that integrate extracted emotion features to improve human value detection tasks. This approach holds the potential to provide fresh insights into the complex interactions between emotions and values within discussions, offering a deeper understanding of human behavior and decision making. Uncovering these emotions is crucial for comprehending the characteristics that underlie various values through data-driven analyses. Our experiment results show improvement in the performance of human value detection tasks in many categories.

Widening Blockchain Technology toward Access Control for Service Provisioning in Cellular Networks.

The attention on blockchain technology (BCT) to create new forms of relational reliance has seen an explosion of new applications and initiatives, to assure decentralized security and trust. Its potential as a game-changing technology relates to how data gets distributed and replicated over several organizations and countries. This paper provides an introduction to BCT, as well as a review of its technological aspects. A concrete application of outsource access control and pricing procedures in cellular networks, based on a decentralized access control-as-a-service solution for private cellular networks, is also presented. The application can be used by service and content providers, to provide new business models. The proposed method removes the single point of failure from conventional centralized access control systems, increasing scalability while decreasing operational complexity, regarding access control and pricing procedures. Design and implementation details of the new method in a real-world scenario using a private cellular network and a BCT system that enables smart contracts are also provided.

Data Analytics, Self-Organization, and Security Provisioning for Smart Monitoring Systems.

Internet availability and its integration with smart technologies have favored everyday objects and things and offered new areas, such as the Internet of Things (IoT). IoT refers to a concept where smart devices or things are connected and create a network. This new area has suffered from big data handling and security issues. There is a need to design a data analytics model by using new 5G technologies, architecture, and a security model. Reliable data communication in the presence of legitimate nodes is always one of the challenges in these networks. Malicious nodes are generating inaccurate information and breach the user's security. In this paper, a data analytics model and self-organizing architecture for IoT networks are proposed to understand the different layers of technologies and processes. The proposed model is designed for smart environmental monitoring systems. This paper also proposes a security model based on an authentication, detection, and prediction mechanism for IoT networks. The proposed model enhances security and protects the network from DoS and DDoS attacks. The proposed model evaluates in terms of accuracy, sensitivity, and specificity by using machine learning algorithms.

Coverage Area Decision Model by Using Unmanned Aerial Vehicles Base Stations for Ad Hoc Networks.

Unmanned Aerial Vehicle (UAV) deployment and placement are largely dependent upon the available energy, feasible scenario, and secure network. The feasible placement of UAV nodes to cover the cellular networks need optimal altitude. The under or over-estimation of nodes' air timing leads to of resource waste or inefficiency of the mission. Multiple factors influence the estimation of air timing, but the majority of the literature concentrates only on flying time. Some other factors also degrade network performance, such as unauthorized access to UAV nodes. In this paper, the UAV coverage issue is considered, and a Coverage Area Decision Model for UAV-BS is proposed. The proposed solution is designed for cellular network coverage by using UAV nodes that are controlled and managed for reallocation, which will be able to change position per requirements. The proposed solution is evaluated and tested in simulation in terms of its performance. The proposed solution achieved better results in terms of placement in the network. The simulation results indicated high performance in terms of high packet delivery, less delay, less overhead, and better malicious node detection.

Improving Road Traffic Forecasting Using Air Pollution and Atmospheric Data: Experiments Based on LSTM Recurrent Neural Networks.

Traffic flow forecasting is one of the most important use cases related to smart cities. In addition to assisting traffic management authorities, traffic forecasting can help drivers to choose the best path to their destinations. Accurate traffic forecasting is a basic requirement for traffic management. We propose a traffic forecasting approach that utilizes air pollution and atmospheric parameters. Air pollution levels are often associated with traffic intensity, and much work is already available in which air pollution has been predicted using road traffic. However, to the best of our knowledge, an attempt to improve forecasting road traffic using air pollution and atmospheric parameters is not yet available in the literature. In our preliminary experiments, we found out the relation between traffic intensity, air pollution, and atmospheric parameters. Therefore, we believe that addition of air pollutants and atmospheric parameters can improve the traffic forecasting. Our method uses air pollution gases, including C O , N O , N O 2 , N O x , and O 3 . We chose these gases because they are associated with road traffic. Some atmospheric parameters, including pressure, temperature, wind direction, and wind speed have also been considered, as these parameters can play an important role in the dispersion of the above-mentioned gases. Data related to traffic flow, air pollution, and the atmosphere were collected from the open data portal of Madrid, Spain. The long short-term memory (LSTM) recurrent neural network (RNN) was used in this paper to perform traffic forecasting.

Variational Channel Estimation with Tempering: An Artificial Intelligence Algorithm for Wireless Intelligent Networks.

With the rapid development of wireless sensor networks (WSNs) technology, a growing number of applications and services need to acquire the states of channels or sensors, especially in order to use these states for monitoring, object tracking, motion detection, etc. A critical issue in WSNs is the ability to estimate the source parameters from the readings of a distributed sensor network. Although there are several studies on channel estimation (CE) algorithms, existing algorithms are all flawed with their high complexity, inability to scale, inability to ensure the convergence to a local optimum, low speed of convergence, etc. In this work, we turn to variational inference (VI) with tempering to solve the channel estimation problem due to its ability to reduce complexity, ability to generalize and scale, and guarantee of local optimum. To the best of our knowledge we are the first to use VI with tempering for advanced channel estimation. The parameters that we consider in the channel estimation problem include pilot signal and channel coefficients, assuming there is orthogonal access between different sensors (or users) and the data fusion center (or receiving center). By formulating the channel estimation problem into a probabilistic graphical model, the proposed Channel Estimation Variational Tempering Inference (CEVTI) approach can estimate the channel coefficient and the transmitted signal in a low-complexity manner while guaranteeing convergence. CEVTI can find out the optimal hyper-parameters of channels with fast convergence rate, and can be applied to the case of code division multiple access (CDMA) and uplink massive multi-input-multi-output (MIMO) easily. Simulations show that CEVTI has higher accuracy than state-of-the-art algorithms under different noise variance and signal-to-noise ratio. Furthermore, the results show that the more parameters are considered in each iteration, the faster the convergence rate and the lower the non-degenerate bit error rate with CEVTI. Analysis shows that CEVTI has satisfying computational complexity, and guarantees a better local optimum. Therefore, the main contribution of the paper is the development of a new efficient, simple and reliable algorithm for channel estimation in WSNs.

A Comparative Analysis of Machine/Deep Learning Models for Parking Space Availability Prediction.

Machine/Deep Learning (ML/DL) techniques have been applied to large data sets in order to extract relevant information and for making predictions. The performance and the outcomes of different ML/DL algorithms may vary depending upon the data sets being used, as well as on the suitability of algorithms to the data and the application domain under consideration. Hence, determining which ML/DL algorithm is most suitable for a specific application domain and its related data sets would be a key advantage. To respond to this need, a comparative analysis of well-known ML/DL techniques, including Multilayer Perceptron, K-Nearest Neighbors, Decision Tree, Random Forest, and Voting Classifier (or the Ensemble Learning Approach) for the prediction of parking space availability has been conducted. This comparison utilized Santander's parking data set, initiated while working on the H2020 WISE-IoT project. The data set was used in order to evaluate the considered algorithms and to determine the one offering the best prediction. The results of this analysis show that, regardless of the data set size, the less complex algorithms like Decision Tree, Random Forest, and KNN outperform complex algorithms such as Multilayer Perceptron, in terms of higher prediction accuracy, while providing comparable information for the prediction of parking space availability. In addition, in this paper, we are providing Top-K parking space recommendations on the basis of distance between current position of vehicles and free parking spots.

Industrial Internet of Things-Based Collaborative Sensing Intelligence: Framework and Research Challenges.

The development of an efficient and cost-effective solution to solve a complex problem (e.g., dynamic detection of toxic gases) is an important research issue in the industrial applications of the Internet of Things (IoT). An industrial intelligent ecosystem enables the collection of massive data from the various devices (e.g., sensor-embedded wireless devices) dynamically collaborating with humans. Effectively collaborative analytics based on the collected massive data from humans and devices is quite essential to improve the efficiency of industrial production/service. In this study, we propose a collaborative sensing intelligence (CSI) framework, combining collaborative intelligence and industrial sensing intelligence. The proposed CSI facilitates the cooperativity of analytics with integrating massive spatio-temporal data from different sources and time points. To deploy the CSI for achieving intelligent and efficient industrial production/service, the key challenges and open issues are discussed, as well.

Sensor Network-Based and User-Friendly User Location Discovery for Future Smart Homes.

User location is crucial context information for future smart homes where many location based services will be proposed. This location necessarily means that User Location Discovery (ULD) will play an important role in future smart homes. Concerns about privacy and the need to carry a mobile or a tag device within a smart home currently make conventional ULD systems uncomfortable for users. Future smart homes will need a ULD system to consider these challenges. This paper addresses the design of such a ULD system for context-aware services in future smart homes stressing the following challenges: (i) users' privacy; (ii) device-/tag-free; and (iii) fault tolerance and accuracy. On the other hand, emerging new technologies, such as the Internet of Things, embedded systems, intelligent devices and machine-to-machine communication, are penetrating into our daily life with more and more sensors available for use in our homes. Considering this opportunity, we propose a ULD system that is capitalizing on the prevalence of sensors for the home while satisfying the aforementioned challenges. The proposed sensor network-based and user-friendly ULD system relies on different types of inexpensive sensors, as well as a context broker with a fuzzy-based decision-maker. The context broker receives context information from different types of sensors and evaluates that data using the fuzzy set theory. We demonstrate the performance of the proposed system by illustrating a use case, utilizing both an analytical model and simulation.

An energy-aware routing protocol for query-based applications in wireless sensor networks.

Wireless sensor network (WSN) typically has energy consumption restriction. Designing energy-aware routing protocol can significantly reduce energy consumption in WSNs. Energy-aware routing protocols can be classified into two categories, energy savers and energy balancers. Energy saving protocols are used to minimize the overall energy consumed by a WSN, while energy balancing protocols attempt to efficiently distribute the consumption of energy throughout the network. In general terms, energy saving protocols are not necessarily good at balancing energy consumption and energy balancing protocols are not always good at reducing energy consumption. In this paper, we propose an energy-aware routing protocol (ERP) for query-based applications in WSNs, which offers a good trade-off between traditional energy balancing and energy saving objectives and supports a soft real time packet delivery. This is achieved by means of fuzzy sets and learning automata techniques along with zonal broadcasting to decrease total energy consumption.

Offensive language detection in low resource languages: A use case of Persian language.

THIS ARTICLE USES WORDS OR LANGUAGE THAT IS CONSIDERED PROFANE, VULGAR, OR OFFENSIVE BY SOME READERS. Different types of abusive content such as offensive language, hate speech, aggression, etc. have become prevalent in social media and many efforts have been dedicated to automatically detect this phenomenon in different resource-rich languages such as English. This is mainly due to the comparative lack of annotated data related to offensive language in low-resource languages, especially the ones spoken in Asian countries. To reduce the vulnerability among social media users from these regions, it is crucial to address the problem of offensive language in such low-resource languages. Hence, we present a new corpus of Persian offensive language consisting of 6,000 out of 520,000 randomly sampled micro-blog posts from X (Twitter) to deal with offensive language detection in Persian as a low-resource language in this area. We introduce a method for creating the corpus and annotating it according to the annotation practices of recent efforts for some benchmark datasets in other languages which results in categorizing offensive language and the target of offense as well. We perform extensive experiments with three classifiers in different levels of annotation with a number of classical Machine Learning (ML), Deep learning (DL), and transformer-based neural networks including monolingual and multilingual pre-trained language models. Furthermore, we propose an ensemble model integrating the aforementioned models to boost the performance of our offensive language detection task. Initial results on single models indicate that SVM trained on character or word n-grams are the best performing models accompanying monolingual transformer-based pre-trained language model ParsBERT in identifying offensive vs non-offensive content, targeted vs untargeted offense, and offensive towards individual or group. In addition, the stacking ensemble model outperforms the single models by a substantial margin, obtaining 5% respective macro F1-score improvement for three levels of annotation.

A survey on multi-lingual offensive language detection.

The prevalence of offensive content on online communication and social media platforms is growing more and more common, which makes its detection difficult, especially in multilingual settings. The term "Offensive Language" encompasses a wide range of expressions, including various forms of hate speech and aggressive content. Therefore, exploring multilingual offensive content, that goes beyond a single language, focus and represents more linguistic diversities and cultural factors. By exploring multilingual offensive content, we can broaden our understanding and effectively combat the widespread global impact of offensive language. This survey examines the existing state of multilingual offensive language detection, including a comprehensive analysis on previous multilingual approaches, and existing datasets, as well as provides resources in the field. We also explore the related community challenges on this task, which include technical, cultural, and linguistic ones, as well as their limitations. Furthermore, in this survey we propose several potential future directions toward more efficient solutions for multilingual offensive language detection, enabling safer digital communication environment worldwide.

Health-ID: A Blockchain-Based Decentralized Identity Management for Remote Healthcare.

COVID-19 has made eHealth an imperative. The pandemic has been a true catalyst for remote eHealth solutions such as teleHealth. Telehealth facilitates care, diagnoses, and treatment remotely, making them more efficient, accessible, and economical. However, they have a centralized identity management system that restricts the interoperability of patient and healthcare provider identification. Thus, creating silos of users that are unable to authenticate themselves beyond their eHealth application's domain. Furthermore, the consumers of remote eHealth applications are forced to trust their service providers completely. They cannot check whether their eHealth service providers adhere to the regulations to ensure the security and privacy of their identity information. Therefore, we present a blockchain-based decentralized identity management system that allows patients and healthcare providers to identify and authenticate themselves transparently and securely across different eHealth domains. Patients and healthcare providers are uniquely identified by their health identifiers (healthIDs). The identity attributes are attested by a healthcare regulator, indexed on the blockchain, and stored by the identity owner. We implemented smart contracts on an Ethereum consortium blockchain to facilities identification and authentication procedures. We further analyze the performance using different metrics, including transaction gas cost, transaction per second, number of blocks lost, and block propagation time. Parameters including block-time, gas-limit, and sealers are adjusted to achieve the optimal performance of our consortium blockchain.

Hate speech detection and racial bias mitigation in social media based on BERT model.

Disparate biases associated with datasets and trained classifiers in hateful and abusive content identification tasks have raised many concerns recently. Although the problem of biased datasets on abusive language detection has been addressed more frequently, biases arising from trained classifiers have not yet been a matter of concern. In this paper, we first introduce a transfer learning approach for hate speech detection based on an existing pre-trained language model called BERT (Bidirectional Encoder Representations from Transformers) and evaluate the proposed model on two publicly available datasets that have been annotated for racism, sexism, hate or offensive content on Twitter. Next, we introduce a bias alleviation mechanism to mitigate the effect of bias in training set during the fine-tuning of our pre-trained BERT-based model for hate speech detection. Toward that end, we use an existing regularization method to reweight input samples, thereby decreasing the effects of high correlated training set' s n-grams with class labels, and then fine-tune our pre-trained BERT-based model with the new re-weighted samples. To evaluate our bias alleviation mechanism, we employed a cross-domain approach in which we use the trained classifiers on the aforementioned datasets to predict the labels of two new datasets from Twitter, AAE-aligned and White-aligned groups, which indicate tweets written in African-American English (AAE) and Standard American English (SAE), respectively. The results show the existence of systematic racial bias in trained classifiers, as they tend to assign tweets written in AAE from AAE-aligned group to negative classes such as racism, sexism, hate, and offensive more often than tweets written in SAE from White-aligned group. However, the racial bias in our classifiers reduces significantly after our bias alleviation mechanism is incorporated. This work could institute the first step towards debiasing hate speech and abusive language detection systems.