Constructing Social Networks

Ensuring privacy of users of social networks is probably an unsolvable conundrum. At the same time, an informed use of the existing privacy options by the social network participants may alleviate -or even prevent -some of the more drastic privacy-averse incidents. Unfortunately, recent surveys show that an average user is either not aware of these options or does not use them, probably due to their perceived complexity. It is therefore reasonable to believe that tools assisting users with two tasks: 1) understanding their social net behaviour in terms of their privacy settings and broad privacy categories, and 2) recommending reasonable privacy options, will be a valuable tool for everyday privacy practice in a social network context. This paper presents YourPrivacyProtector, a recommender system that shows how simple machine learning techniques may provide useful assistance in these two tasks to Facebook users. We support our claim with empirical results of application of YourPrivacyProtector to two groups of Facebook users.

Ensuring privacy of users of social networks is probably an unsolvable conundrum. At the same time, an informed use of the existing privacy options by the social network participants may alleviate -or even prevent -some of the more drastic privacy-averse incidents. Unfortunately, recent surveys show that an average user is either not aware of these options or does not use them, probably due to their perceived complexity. It is therefore reasonable to believe that tools assisting users with two tasks: 1) understanding their social net behaviour in terms of their privacy settings and broad privacy categories, and 2) recommending reasonable privacy options, will be a valuable tool for everyday privacy practice in a social network context. This paper presents YourPrivacyProtector, a recommender system that shows how simple machine learning techniques may provide useful assistance in these two tasks to Facebook users. We support our claim with empirical results of application of YourPrivacyProtector to two groups of Facebook users.

ICT is widely adopted by Asian youth and is utilized by people of all ages across the continent. Despite its many advantages, unethical ICT usage can lead to many complications. A harmful application of ICT for social communication and engagement is cyberbullying. Simply adhering to the generally accepted norms and guidelines for cybersecurity will not protect you from cybercrime. Even well-known social media stages like Twitter are safe from this attack. Natural language processing (NLP) research on cyberbullying detection has become popular recently. Even though old-style NLP procedures have become highly cyberbullying, there are still hurdles to overcome. These include the limited character count allowed by social media platforms, an imbalance among comments, ambiguity, and unnecessary use of slang. Models based on (CNNs), Multilayer Perceptrons (MLPs), and (RNNs), have recently shown encouraging results in a variety of NLP tasks. With this motivation, this research develops an A...

The rapid growth of social media in recent years has fed into some highly undesirable phenomena such as proliferation of abusive and offensive language on the Internet. Previous research suggests that such hateful content tends to come from users who share a set of common stereotypes and form communities around them. The current state-of-the-art approaches to hate speech detection are oblivious to user and community information and rely entirely on textual (i.e., lexical and semantic) cues. In this paper, we propose a novel approach to this problem that incorporates community-based profiling features of Twitter users. Experimenting with a dataset of 16k tweets, we show that our methods significantly outperform the current state of the art in hate speech detection. Further, we conduct a qualitative analysis of model characteristics. We release our code, pre-trained models and all the resources used in the public domain.

The rapid growth of social media in recent years has fed into some highly undesirable phenomena such as proliferation of hateful and offensive language on the Internet. Previous research suggests that such abusive content tends to come from users who share a set of common stereotypes and form communities around them. The current state-of-the-art approaches to abuse detection are oblivious to user and community information and rely entirely on textual (i.e., lexical and semantic) cues. In this paper, we propose a novel approach to this problem that incorporates community-based profiling features of Twitter users. Experimenting with a dataset of 16k tweets, we show that our methods significantly outperform the current state of the art in abuse detection. Further, we conduct a qualitative analysis of model characteristics. We release our code, pre-trained models and all the resources used in the public domain.

The article is devoted to determining the types of youth content in social networks produced by political power subjects. The study's empirical base is the accounts of the social network Vkontakte, which are politically oriented towards young people and with the number of participants from six thousand. The statistical data analysis identified types of content and their typology according to the indicator of social engagement. Content monitoring of political posts, as well as obtaining basic statistics on social engagement, allowed to identify five main tasks of youth political Internet communication: increasing community members, identifying the needs of the target audience, building the reputation of the organization and the image of a political leader, engaging the target audience, campaigning, and also offer a set of types and forms of content for their practical implementation.

Automatic Text Classification is a machine learning task that automatically assigns a given text document to a set of pre-defined categories based on the features extracted from its textual content. Most online communication forums, including social media, enable users to express themselves freely, and most times, anonymously. The ability to freely express oneself is a human right that should be cherished, but people always induce and spread hate or illegal words towards another group as an abuse of this liberty. For instance many online forums such as Facebook, YouTube, and Twitter consider hate speech harmful, and have policies to remove hate speech content. This paper attempts to automatically classify the textual entries made by bloggers on various topics into hate speech and non-hate speech. This was achieved by following steps like pre-processing, feature extraction and support vector machine classification. Empirical evaluation of this binary classification has resulted in anaccuracy of approximately 83% over the test set. In addition to classifying the textual entries of the blogs, it is proposed that the extracted features themselves be further classified under more meaningful heads which results in generation of a semantic resource that lends greater understanding to the classification task. This semantic resource can be used for data mining requirements that arise in the future.

The Internet has revolutionized the way Americans process information. Social media platforms such as Twitter and Facebook have enabled discourse among citizens in a way never before possible. Often, these discussions are largely driven by partisan divide, where individuals sway towards predetermined schools of thought. Normally, it is healthy to discuss current events and analyze perspectives from all sides. However, social media has also allowed people to engage negatively with others and has provided a platform for hate speech. As the country has become increasingly polarized over certain issues, hate speech as become all the more common. Given the relationship between politicallycharged rhetoric and hate speech, we seek to determine whether this relationship can be observed through transfer learning. In this paper, we study the effect of pre-training on partisan/non-partisan political tweets for a hate speech classification model. We hypothesize that pre-trained parameters from ...

The widespread use of social media has escalated concerns about cyberbullying. Traditional methods for detecting and managing
cyberbullying struggle with the sheer volume of electronic text data. This has led to the exploration of deep learning as a potential
solution. Researchers study focuses on implementing deep learning techniques to identify cyberbullying in Arabic social media,
specifically targeting three prevalent forms of Arabic: dialectal, Modern Standard, and Classical. The collected data corpus was
about 30, 0000 tweets. In this work, we first examined the sentiment analysis as cyberbullying, and No cyberbullying, then we
further classified the cyberbullying by labelling the data under six different cyberbullying categories. We implemented deep
learning models such as CNN, RNN, and a combination of CNN-RNN. The results that obtained from 2-classes classification
showed a superiority of LSTM in terms of accuracy with 95.59%, while the best accuracy in the 6-classes classification gained
from implementing CNN with 78.75%. Meanwhile the f1-score results were the highest in LSTM for the 2-lasses and 6-classes
classifications with 96.73%, and 89%, respectively. These findings emphasize the potential for deep learning techniques to be
applied in the development of automated systems for identifying and combating cyberbullying on social media and show how well
they work in detecting cyberbullying.

How well can hate speech concept be abstracted in order to inform automatic classification in codeswitched texts by machine learning classifiers? We explore different representations and empirically evaluate their predictiveness using both conventional and deep learning algorithms in identifying hate speech in a ~48k human-annotated dataset that contain mixed languages, a phenomenon common among multilingual speakers. This paper espouses a novel approach to handle this challenge by introducing a hierarchical approach that employs Latent Dirichlet Allocation to generate topic models that feed into another high-level feature set that we acronym PDC. PDC groups similar meaning words in word families during the preprocessing stage for supervised learning models. The high-level PDC features generated are based on Ombui et al, (2019) hate speech annotation framework that is informed by the triangular theory of hate (Stanberg,2003). Results obtained from frequency-based models using the PD...

How well can hate speech concept be abstracted in order to inform automatic classification in codeswitched texts by machine learning classifiers? We explore different representations and empirically evaluate their predictiveness using both conventional and deep learning algorithms in identifying hate speech in a ~48k human-annotated dataset that contain mixed languages, a phenomenon common among multilingual speakers. This paper espouses a novel approach to handle this challenge by introducing a hierarchical approach that employs Latent Dirichlet Allocation to generate topic models that feed into another high-level feature set that we acronym PDC. PDC groups similar meaning words in word families during the preprocessing stage for supervised learning models. The high-level PDC features generated are based on Ombui et al, (2019) hate speech annotation framework that is informed by the triangular theory of hate (Stanberg,2003). Results obtained from frequency-based models using the PD...

People feel entitled to express their opinions on social media, and many of these opinions may be hostile to others. Hate speech recognition is essential in today's world, where it is important to keep track of what is posted in public for all to see. This paper discusses ways to detect hateful language on any topic using data mining, natural language processing, and machine learning techniques on the social media site Twitter. A Logistic Regression Model for Hate Speech Recognition is proposed. The research conducted provides details of all the hate commenters as well as an overview of the topics on which the hate is being projected the most.

The wealth of opinions expressed in micro-blogs, such as tweets, motivated researchers to develop techniques for automatic opinion detection. However, accuracies of such techniques are still limited. Moreover, current techniques focus on detecting sentiment polarity regardless of the topic (target) discussed. Detecting sentiment towards a specific target, referred to as target-dependent sentiment classification, has not received adequate researchers’ attention. Literature review has shown that all target-dependent approaches use supervised learning techniques. Such techniques need a large number of labeled data. However, labeling data in social media is cumbersome and error prone. The research presented in this paper addresses this issue by employing semi-supervised learning techniques for target-dependent sentiment classification. Semisupervised learning techniques make use of labeled as well as unlabeled data. In this paper, we present a new semi-supervised learning technique that...

The advancement of social media plays an important role in increasing the population of youngsters on the web. And it has become the biggest medium of expressing one's thoughts and emotions. Recent studies report that cyberbullying constitutes a growing problem among youngsters on the web. These kinds of attacks have a major influence on the current generation's personal and social life because youngsters are ready to adopt online life instead of a real one, which leads them into an imaginary world. So, we are proposing a system for early detection of cyberbullying on the web and comparing different machine learning Algorithms to obtain the optimal result. We are comparing four different algorithms which can be effectively used for the detection of cyberbullying, with the implementation of the bag of words algorithm with different n-gram methods. Comparatively naïve Bayes algorithm has the highest accuracy of 79% with trigram implementation of the bag of words algorithm.

Social media has affected people's information sources. Since most of the news on social media is not verified by a central authority, it may contain fake news for various reasons such as advertising and propaganda. Considering an average of 500 million tweets were posted daily on Twitter alone in the year of 2020, it is possible to control each share only with smart systems. In this study, we use Natural Language Processing methods to detect fake news for Turkish-language posts on certain topics on Twitter. Furthermore, we examine the follow/follower relations of the users who shared fake-real news on the same subjects through social network analysis methods and visualization tools. Various supervised and unsupervised learning algorithms have been tested with different parameters. The most successful F1 score of fake news detection was obtained with the support vector machines algorithm with 0.9. People who share fake/true news can help in the separation of subgroups in the social network created by people and their followers. The results show that fake news propagation networks may show different characteristics in their own subject based on the follow/follower network.

Antisocial behavior is a personality disorder that has characteristics such as repetitive actions that violate social norms, deceit and lying, impulsiveness, irritability and aggression, reckless disregard for the safety of oneself and others, consistently irresponsible, and lack of remorse. The cause can be from various factors, including genetics, psychological conditions, interactions in the environment, and wrong parenting. The impact of antisocial behavior on social life can cause people to tend to be aggressive and take it into action by not having feelings of guilt for their actions. Thus, a monitoring of antisocial behavior disorders is needed so that it can be a warning for the public to be more concerned about the difficulties experienced by each other. The potential gained from the availability of tweet data access from the Twitter API opens up opportunities for monitoring antisocial behavior. By utilizing traditional machine learning and deep learning methods, it can be ...

Since the textual contents on online social media are highly unstructured, informal, and often misspelled, existing research on message-level offensive language detection cannot accurately detect offensive content. This feature of the social media to express something openly to the world have created the major problems for these online businesses and negatively impacted the well-being of the societal decorum. There are increasing cases of the abuse or offense on the social media like Hate speech, Cyber-bullying, Aggression or general Profanity. It is very much important to understand that this behavior can not only immensely affect the life of an individual or a group but could be suicidal in some cases adversely hampering the mental health of the victims .

Controversies are an inseparable part of social systems which, if constructed properly, can create a unique condition for higher-order learning. In addition, design inquiry, as a process of thought and planning, is also a constructive process. This provokes the question of how to construct controversies from a designerly perspective in order to steer higher-order learning. This paper presents a theoretical contribution to the field of social system design by providing the first insights into design intervention to facilitate a network of allied construction. Through a systematic review of the concept of conflict and disagreement, the link between controversies and knowledge transmission is examined in order to highlight the benefit of controversies in a constructive way. Next to that, the essential steps for constructing a network of allies are proposed. These steps are compared with specific aspects of design in order to unfold the advantages of design for network construction. Fin...

Twitter is a social media platform and has become a go to place for young and old to express their views in short digital text format. This has made Twitter a source of big data of millions of small tweets.
This mini project tries to survey machine learning techniques for hate speech classification from Twitter datasets available for public access. Hate speech classification on Twitter datasets remains a vibrant research focus, but little research efforts have been devoted to the design of a generic metadata architecture, threshold settings and fragmentation issues. Hate speech classification techniques presented in literature address some of the challenges inherent in Twitter
data streams but limited in the aforementioned issues.
This project presented collection of hate speech benchmarks datasets suitable for testing the efficiency of classification models. This project also presented the pros and cons for single and hybrid machine learning methods in hate speech classification. The summary of performance evaluation for the surveyed machine learning methods was also presented.
The project also presented a generic metadata architecture for hate speech classification in Twitter to tackle issues with Twitter datasets. The developed generic metadata architecture was observed to performed better across all evaluation metrics for hate speech detection having 0.95, 0.93, 0.92 and 0.93 for accuracy, precision, recall and F1-score respectively, when compared to similar methods. Similarly, the developed generic metadata architecture for hate speech sentiment
classification performed better with F1-score of 91.5% compared to related methods.
The developed generic metadata architecture also indicates a more perfect test having an AUC of 0.97, when compared to similar methods. The statistical validation of results points out the efficiency of the developed system.
Finally, the results also showed that the developed system is very good for automatic topic detection and categorization.

There are growing signs of discontent with the antisocial behavior expressed on social media platforms. Harnessing the power of machine learning for the purpose of detecting and mediating the spread of malicious behavior has received a heightened attention in the last decade. In this paper, we report on an experiment that examines the predictive power of a number of sparse and dense feature representations coupled with a multi-level ensemble classifier. To address the research questions, we have used PAN 2021 Profiling Hate Speech Spreaders on Twitter task for English language. The initial results are encouraging pointing out to the robustness of the proposed model when evaluated on the test dataset.

Sentiment Analysis, often known as opinion mining, is the technique of determining whether text data is positive, neutral, or negative using natural language processing, computer linguistics, and related technology. Hate Speech, on the other hand, is abusive or threatening speech or writing that conveys prejudice, hatred, and/or urges violence against a person or group of people because of their race, religion, sex, or sexual orientation.With the exponential rise in the number of individuals using social media, where lots of stuff is shared everyday that is ostensibly harmless, hate speech has also increased dramatically. The necessity for identifying and detecting hate speech on social media platforms, particularly Twitter, has grown considerably, as large corporations work to establish mechanisms to combat hate speech online. The goal of this project is to meet the demand for recognising unfavourable tweets that promote hate speech. 3

Sentiment analysis mainly supports sorting out the polarity and provides valuable information with the use of raw data in social media platforms. Many fields like health, business, and security require real-time data analysis for instant decision-making situations.Since Twitter is considered a popular social media platform to collect data easily, this paper is considering data analysis methods of Twitter data, real-time Twitter data analysis based on geo-location. Twitter data classification and analysis can be done with the use of diverse algorithms and deciding the most appropriate algorithm for data analysis, can be accomplished by implementing and testing these diverse algorithms.This paper is discussing the major description of sentiment analysis, data collection methods, data pre-processing, feature extraction, and sentiment analysis methods related to Twitter data. Real-time data analysis arises as a major method of analyzing the data available online and the real-time Twitte...

Social networks have ensured the expanding disproportion between the face of WWW stored traditionally in search engine repositories and the actual ever changing face of Web. Exponential growth of web users and the ease with which they can upload contents on web highlights the need of content controls on material published on the web. As definition of search is changing, socially-enhanced interactive search methodologies are the need of the hour. Ranking is pivotal for efficient web search as the search performance mainly depends upon the ranking results. In this paper new integrated ranking model based on fused rank of web object based on popularity factor earned over only valid interlinks from multiple social forums is proposed. This model identifies relationships between web objects in separate social networks based on the object inheritance graph. Experimental study indicates the effectiveness of proposed Fusion based ranking algorithm in terms of better search results.

Hate-speech detection on social network language has become one of the main researching fields recently due to the spreading of social networks like Facebook and Twitter. In Vietnam, the threat of offensive and harassment cause bad impacts for online user. The VLSP-Shared task about Hate Speech Detection on social networks showed many proposed approaches for detecting whatever comment is clean or not. However, this problem still needs further researching. Consequently, we compare traditional machine learning and deep learning on a large dataset about the user's comments on social network in Vietnamese and find out what is the advantage and disadvantage of each model by comparing their accuracy on F1score, then we pick two models in which has highest accuracy in traditional machine learning models and deep neural models respectively. Next, we compare these two models capable of predicting the right label by referencing their confusion matrices and considering the advantages and disadvantages of each model. Finally, from the comparison result, we propose our ensemble method that concentrates the abilities of traditional methods and deep learning methods.

Socialiniai tinklai-greitai populiarėjanti informacijos sklaidos priemonė. Pastebima, kad socialiniai tinklai paaugliams yra traukos vieta. Dažnas ir ilgas paauglių prisijungimas prie socialinių tinklų skatina priklausomybę. Tyrimo tikslas-atskleisti paauglių, priklausomų nuo socialinių tinklų, naudojimosi jais ir elgesio ypatumus. Atliekant tyrimą naudotas kiekybinis aprašomasis metodas, anketinė apklausa raštu. Atliekant tyrimą atskleista, kad paaugliams būdingas priklausomas nuo socialinių tinklų elgesys. Nors daugelis paauglių teigia, kad nėra priklausomi nuo socialinių tinklų, tačiau gana stipriai pasireiškia priklausomas paauglių elgesys nuo jų psichologiniu, socialiniu ir kiek mažiau fiziologiniu požiūriu.

Vytauto Didžiojo universitetasK. Donelaičio g. 58 LT-44248 KaunasTel. (8 37) 32 78 20El. paštas: [email protected] tinklai – greitai populiarėjanti informacijos sklaidos priemonė. Pastebima, kad socialiniai tinklai paaugliams yra traukos vieta. Dažnas ir ilgas paauglių prisijungimas prie socialinių tinklų skatina priklausomybę. Tyrimo tikslas – atskleisti paauglių, priklausomų nuo socialinių tinklų, naudojimosi jais ir elgesio ypatumus. Atliekant tyrimą naudotas kiekybinis aprašomasis metodas, anketinė apklausa raštu. Atliekant tyrimą atskleista, kad paaugliams būdingas priklausomas nuo socialinių tinklų elgesys. Nors daugelis paauglių teigia, kad nėra priklausomi nuo socialinių tinklų, tačiau gana stipriai pasireiškia priklausomas paauglių elgesys nuo jų psichologiniu, socialiniu ir kiek mažiau fiziologiniu požiūriu.Pagrindinės sąvokos: paaugliai, priklausomybė, socialiniai tinklai internetinėje erdvėje.Teenagers’ Addiction to Social Networks on the InternetTomas Bac...

Sentiment Analysis, often known as opinion mining, is the technique of determining whether text data is positive, neutral, or negative using natural language processing, computer linguistics, and related technology. Hate Speech, on the other hand, is abusive or threatening speech or writing that conveys prejudice, hatred, and/or urges violence against a person or group of people because of their race, religion, sex, or sexual orientation.With the exponential rise in the number of individuals using social media, where lots of stuff is shared everyday that is ostensibly harmless, hate speech has also increased dramatically. The necessity for identifying and detecting hate speech on social media platforms, particularly Twitter, has grown considerably, as large corporations work to establish mechanisms to combat hate speech online. The goal of this project is to meet the demand for recognising unfavourable tweets that promote hate speech. 3

Abstract: Cyberbullying or cyber harassment is a form of bullying or harassment using electronic means. Cyberbullying and cyber harassment are also known as online bullying. It has become increasingly common, especially among teenagers, as the digital sphere has expanded and technology has advanced. In this research, we have addressed the problem of cyberbullying detection on Twitter Data Set. Various inbuilt Multiclass Classification Algorithms in Python such as Naive Bayes, Decision tree, Random Forest, K Nearest Neighbor(KNN), Support Vector Machine (SVM), and Natural Language Processing(NLP) techniques are used to classify bullying and non-bullying tweets and performance of these algorithms is compared and model with highest accuracy is selected for the prediction. Bullying tweets will be reported to the nearest Cyber Crime Branch.

Today, the internet is an integral part of our daily lives, enabling people to be more connected than ever before. However, this greater connectivity and access to information increase exposure to harmful content such as cyber-bullying and cyber-hatred. Models based on machine learning and natural language offer a way to make online platforms safer by identifying hate speech in web text autonomously. However, the main difficulty is annotating a sufficiently large number of examples to train these models. This paper uses a transfer learning technique to leverage two independent datasets jointly and builds a single representation of hate speech. We build an interpretable two-dimensional visualization tool of the constructed hate speech representation-dubbed the Map of Hate-in which multiple datasets can be projected and comparatively analyzed. The hateful content is annotated differently across the two datasets (racist and sexist in one dataset, hateful and offensive in another). However, the common representation successfully projects the harmless class of both datasets into the same space and can be used to uncover labeling errors (false positives). We also show that the joint representation boosts prediction performances when only a limited amount of supervision is available. These methods and insights 1

Half of the entire social media users in Indonesia has experienced cyberbullying. Cyberbullying is one of the treatments received as an attack with abusive words. An abusive word is a word or phrase that contained harassment and is expressed be it spoken or in the form of text. This is a serious problem that must be controlled because the act has an impact on the victim's psychology and causes trauma resulting in depression. This study proposed to identify abusive comments from social media in Indonesian language using a deep learning approach. The architecture used is a hybrid model, a combination between recurrent neural network (RNN) and long short-term memory (LSTM). RNN can map the input sequences to fixed-size vectors on hidden vector components and LSTM implemented to overcome gradient vector growth components that have the potential to exist in RNN. The steps carried out include preprocessing, modelling, implementation, and evaluation. The dataset used is indonesian abusive and hate speech from Twitter data. The evaluation result showed that the model proposed produced an f-measure value of 94% with an increase in accuracy of 23%.

How well can hate speech concept be abstracted in order to inform automatic classification in codeswitched texts by machine learning classifiers? We explore different representations and empirically evaluate their predictiveness using both conventional and deep learning algorithms in identifying hate speech in a ~48k human-annotated dataset that contain mixed languages, a phenomenon common among multilingual speakers. This paper espouses a novel approach to handle this challenge by introducing a hierarchical approach that employs Latent Dirichlet Allocation to generate topic models that feed into another high-level feature set that we acronym PDC. PDC groups similar meaning words in word families during the preprocessing stage for supervised learning models. The high-level PDC features generated are based on Ombui et al, (2019) hate speech annotation framework that is informed by the triangular theory of hate (Stanberg,2003). Results obtained from frequency-based models using the PD...

Akivaizdu, kad naudojimasis socialiniais tinklalapiais nuosekliai auga. Lietuvoje pastaraisiais metais vyksta spartus naudojimasis, įskaitant ir vietinius (nacionalinius) socialinius tinklus. "Facebook" yra tarp 10 labiausiai lankomų tinklalapių Lietuvoje. Nemaža dalis Lietuvos vartotojų naudojasi keletu socialinių tinklalapių. Tai susiję su įvairiomis visuomenės veiklos sritimis. Socialiniai tinklai "keliasi" ir į edukacinę erdvę. Dabartiniu metu daugybė įvairių socialinių tinklalapių (ar technologijų, skirtų socialiniams tinklams kurti ir juos efektyviai išnaudoti) naudojami ne tik pramogoms, laisvalaikiui ar bendravimui, bet ir mokymo / mokymosi tikslais. Ar gali socialiniai tinklai efektyviai atlikti edukacinę funkciją? Ar gali jie tapti tinkamu profesinio tobulėjimo instrumentu? Galime formuluoti daugybę klausimų. Atliktas Lietuvos universitetų studentų tyrimas apie jų naudojimąsi socialiniais tinklais atskleidžia potencialias jų panaudojimo galimybes, taip pat kai kuriuos probleminius momentus. Nustatyta, kad nors respondentai žino gana daug socialinių tinklalapių, dažniausiai lankomi ir populiariausi yra trys: "You Tube", "Facebook", "Google+". Tačiau edukacinis socialinių tinklalapių potencialas studentų vertinimuose išreikštas gana silpnai ir tai kelia pagrįstą susirūpinimą.

Hate speech is a prevalent issue on social media platforms that has recently been a cause for concern due to its detrimental effects on individuals and society. The development of effective hate speech detection procedures and algorithms is crucial to address this issue. However, the existing natural language processing (NLP) algorithms and machine learning models face several challenges in accurately identifying and categorizing hate speech. These challenges include the ambiguity and variability of language use, the lack of standardized definitions and guidelines for hate speech, and the rapid evolution of new and creative forms of hate speech. In this paper, we propose a technique that leverages classic machine learning and deep learning methods to locate and categorize hate speech in social media. Our approach involves the use of Support Vector Machines (SVM) and Long Short-Term Memory (LSTM) networks for classification. We evaluate the performance of our model on a hate speech dataset and compare it with a deep learning-based model. Our results show that the SVM model outperforms the deep learning-based model in accuracy and efficiency. Our approach offers a promising solution to the challenges posed by hate speech detection on social media and contributes towards building a safer and more welcoming online community.

Businesses are more concerned than ever about hate speech content as most brand communication and advertising move online. Different organisations may be incharge of their products and services but they do not have complete control over their content posted online via their website and social media channels, they have no control over what online users post or comment about their brand. As a result, it became imperative in our study to develop a model that will identify hate speechand, offensive language and detect cyber offence in online published content using machine learning. This study employed an experimental design to develop a detection model for determining which agile methodologies were preferred as a suitable development methodology. Deep learning and HateSonar was used to detect hate speech and offensive language in posted content. This study used data from Twitter and Facebook to detect hate speech. The text was classified as either hate speech, offensive language, or bo...

Social media has shown to be a medium for the exchange of information, conducting commerce, and even for religious and political activity. The same site has been used to disseminate cyberbullying and hate speech. Cyberbullying is an offense when a perpetrator targets a victim with online provocation and resentment, which has negative impacts on the victim's emotions, relationships, and physical health. These spew forth a lot of hate speech directed at those who have different beliefs or hold different opinions. In order to overcome these problems, this research creates a deeper neural network and maximum entropy-based model for the identification of cyberbullying. In comparison to the present systems, Convolution Neural Network is used for the superior results. The dataset gathered includes 24,783 records tweets with the categories "bullying language," "non-bullying language," and "neither" are included. According to geopolitical zones, tweets are c...

Social media has shown to be a medium for the exchange of information, conducting commerce, and even for religious and political activity. The same site has been used to disseminate cyberbullying and hate speech. Cyberbullying is an offense when a perpetrator targets a victim with online provocation and resentment, which has negative impacts on the victim's emotions, relationships, and physical health. These spew forth a lot of hate speech directed at those who have different beliefs or hold different opinions. In order to overcome these problems, this research creates a deeper neural network and maximum entropy-based model for the identification of cyberbullying. In comparison to the present systems, Convolution Neural Network is used for the superior results. The dataset gathered includes 24,783 records tweets with the categories "bullying language," "non-bullying language," and "neither" are included. According to geopolitical zones, tweets are c...

Cyberbullying (CB) is classified as one of the severe misconducts on social media. Many CB detection systems have been developed for many natural languages to face this phenomenon. However, Arabic is one of the under-resourced languages suffering from the lack of quality datasets in many computational research areas. This paper discusses the design, construction, and evaluation of a multi-dialect, annotated Arabic Cyberbullying Corpus (ArCybC), a valuable resource for Arabic CB detection and motivation for future research directions in Arabic Natural Language Processing (NLP). The study describes the phases of ArCybC compilation. By way of illustration, it explores the corpus to discover strategies used in rendering Arabic CB tweets pulled from four Twitter groups, including gaming, sports, news, and celebrities. Based on thorough analysis, we discovered that these groups were the most susceptible to harassment and cyberbullying. The collected tweets were filtered based on a compiled harassment lexicon, which contains a list of multi-dialectical profane words in Arabic compiled from four categories: sexual, racial, physical appearance, and intelligence. To annotate ArCybC, we asked five annotators to classify 4,505 tweets into two classes manually: Offensive/non-Offensive and CB/non-CB. We conducted a rigorous comparison of different machine learning approaches applied on ArCybC to detect Arabic CB using two language models: bag-of-words (BoW) and word embedding. The experiments showed that Support Vector Machine (SVM) with word embedding achieved an accuracy rate of 86.3% and an F1-score rate of 85%. The main challenges encountered during the ArCybC construction were the scarcity of freely available Arabic CB texts and the deficiency of annotating the texts.

Cyberbullying (CB) is classified as one of the severe misconducts on social media. Many CB detection systems have been developed for many natural languages to face this phenomenon. However, Arabic is one of the under-resourced languages suffering from the lack of quality datasets in many computational research areas. This paper discusses the design, construction, and evaluation of a multi-dialect, annotated Arabic Cyberbullying Corpus (ArCybC), a valuable resource for Arabic CB detection and motivation for future research directions in Arabic Natural Language Processing (NLP). The study describes the phases of ArCybC compilation. By way of illustration, it explores the corpus to discover strategies used in rendering Arabic CB tweets pulled from four Twitter groups, including gaming, sports, news, and celebrities. Based on thorough analysis, we discovered that these groups were the most susceptible to harassment and cyberbullying. The collected tweets were filtered based on a compiled harassment lexicon, which contains a list of multi-dialectical profane words in Arabic compiled from four categories: sexual, racial, physical appearance, and intelligence. To annotate ArCybC, we asked five annotators to classify 4,505 tweets into two classes manually: Offensive/non-Offensive and CB/non-CB. We conducted a rigorous comparison of different machine learning approaches applied on ArCybC to detect Arabic CB using two language models: bag-of-words (BoW) and word embedding. The experiments showed that Support Vector Machine (SVM) with word embedding achieved an accuracy rate of 86.3% and an F1-score rate of 85%. The main challenges encountered during the ArCybC construction were the scarcity of freely available Arabic CB texts and the deficiency of annotating the texts.

There has been a huge growth in recent years interest in studies on abusive language and cyberbullying detection due to its effects on both individual victims and societies. Hate speech, bullying, racism, aggressive content, harassment and other forms of abuse have all significantly increased as a result of Facebook, Instagram, and other social media platforms (SMPs). Since there is a significant need to detect, control, and prohibit the circulation of offensive content on social networking sites, we undertook this study to automate the identification of abusive language or cyberbullying. Arabic data set is balanced and will be used in the offensive detection process. Recently, ensemble machine learning has been used to increase the effectiveness of categorization models. Arabic detection is more precise given that each spatial feature text can make references to every other contextual piece of information. The authors utilized a model that merged convolutional neural network (CNN) with bidirectional long short-term memory (Bi-LSTM) and inverse document frequency gated recurrent unit (GRU) in a hybrid fashion without any post-processing. Our work outperformed every other publicly released cutting-edge ensemble model in the specifications of the official deep learning challenge. The findings indicate that the three-layer inverse document frequency long short-term memory (LSTM) classifier surpassed other classifiers in accuracy with a score of 92.75% compared to different algorithms.

Mining social web text has been at the heart of the Natural Language Processing and Data Mining research community in the last 15 years. Though most of the reported work is on widely spoken languages, such as English, the significance of approaches that deal with less commonly spoken languages, such as Greek, is evident for reasons of preserving and documenting minority languages, cultural and ethnic diversity, and identifying intercultural similarities and differences. The present work aims at identifying, documenting and comparing social text data sets, as well as mining techniques and applications on social web text that target Modern Greek, focusing on the arising challenges and the potential for future research in the specific less widely spoken language.

The use of social media has grown exponentially over time with the growth of the Internet and has become the most influential networking platform in the 21st century. However, the enhancement of social connectivity often creates negative impacts on society that contribute to a couple of bad phenomena such as online abuse, harassment, cyberbullying, cybercrime, and online trolling. Cyberbullying frequently leads to serious mental and physical distress, particularly for women and children, and sometimes even forces them to attempt suicide. Online harassment attracts attention due to its strong negative social impact. Many incidents have recently occurred worldwide due to online harassment, such as sharing private chats, rumours, and sexual remarks. Therefore, the identification of bullying texts or messages on social media has gained a growing amount of attention among researchers. This research aims to design and develop an effective technique to detect online abusive and bullying messages by merging convolutional neural networks (CNN) and deep learning. Six distinct features, namely bagof-words (BoW) and term frequency-inverse text frequency (TFIDF), ngrams, sentiment scores, finding profanity words, and counting emojis, are used to analyse the accuracy level of the deep learning algorithm.

This paper describes SalamNET, an Arabic offensive language detection system that has been submitted to SemEval 2020 shared task 12: Multilingual Offensive Language Identification in Social Media. Our approach focuses on applying multiple deep learning models and conducting in depth error analysis of results to provide system implications for future development considerations. To pursue our goal, a Recurrent Neural Network (RNN), a Gated Recurrent Unit (GRU), and Long-Short Term Memory (LSTM) models with different design architectures have been developed and evaluated. The SalamNET, a Bi-directional Gated Recurrent Unit (Bi-GRU) based model, reports a macro-F1 score of 0.83.

Businesses are more concerned than ever about hate speech content as most brand communication and advertising move online. Different organisations may be incharge of their products and services but they do not have complete control over their content posted online via their website and social media channels, they have no control over what online users post or comment about their brand. As a result, it became imperative in our study to develop a model that will identify hate speechand, offensive language and detect cyber offence in online published content using machine learning. This study employed an experimental design to develop a detection model for determining which agile methodologies were preferred as a suitable development methodology. Deep learning and HateSonar was used to detect hate speech and offensive language in posted content. This study used data from Twitter and Facebook to detect hate speech. The text was classified as either hate speech, offensive language, or bo...

People crave interaction and connection with other people. Therefore, social media became the center of society's life. Among the brightest social media platforms nowadays with a massive number of daily users there is Instagram, which is due to its distinctive features. The excessive revealing of personal life has put users in the spots of getting bullied and harassed and getting toxic revues from other users. Numerous studies have targeted social media to fight its harmful side effects. Nevertheless, most of the datasets that were already available were in English, the Arabic Moroccan dialect ones were not. In this work, the Arabic Moroccan dialect dataset has been extracted from the Instagram platform. Furthermore, feature extraction techniques have been applied to the collected dataset to increase classification accuracy. Afterward, we developed models using machine learning and deep learning algorithms to detect and classify toxicity. For the models' evaluation, we have used the most used metrics: accuracy, precision, F1-score, and recall. The experimental results gave modest scores of around 70% to 83%. These results imply that the models need improvement due to the lack of available datasets and the preprocessing libraries to handle the Moroccan dialect of Arabic.

There is growing interest in automating crime detection and prevention for large populations as a result of the increased usage of social media for victimization and criminal activities. This area is frequently researched due to its potential for enabling criminals to reach a large audience. While several studies have investigated specific crimes on social media, a comprehensive review paper that examines all types of social media crimes, their similarities, and detection methods is still lacking. The identification of similarities among crimes and detection methods can facilitate knowledge and data transfer across domains. The goal of this study is to collect a library of social media crimes and establish their connections using a crime taxonomy. The survey also identifies publicly accessible datasets and offers areas for additional study in this area.