Izzat Alsmadi

Yarmouk University, Computer Information Systems, Faculty Member

Followers

203

Following

Co-authors

Public Views

Izzat Alsmadi is an Associate Professor in the department of computing and cyber security at the Texas A

less

Interests

Uploads

Papers by Izzat Alsmadi

Fake or not? Automated detection of COVID-19 misinformation and disinformation in social networks and digital media

Computational and Mathematical Organization Theory

With the continuous spread of the COVID-19 pandemic, misinformation poses serious threats and con... more With the continuous spread of the COVID-19 pandemic, misinformation poses serious threats and concerns. COVID-19-related misinformation integrates a mixture of health aspects along with news and political misinformation. This mixture complicates the ability to judge whether a claim related to COVID-19 is information, misinformation, or disinformation. With no standard terminology in information and disinformation, integrating different datasets and using existing classification models can be impractical. To deal with these issues, we aggregated several CO-VID-19 misinformation datasets and compared differences between learning models from individual datasets versus one that was aggregated. We also evaluated the impact of using several word-and sentence-embedding models and transformers on the performance of classification models. We observed that whereas word-embedding models showed improvements in all evaluated classification models, the improvement level varied among the different classifiers. Although our work was focused on COVID-19 misinformation detection, a similar approach can be applied to myriad other topics, such as the recent Russian invasion of Ukraine.

Download

Toward Autonomous and Collaborative Information-Credibility-Assessment Systems

Procedia Computer Science, 2020

Online social networks (OSNs) typically promote information sharing and publicity without conside... more Online social networks (OSNs) typically promote information sharing and publicity without consideration of the nature of the content's credibility and quality issues in created and shared content. OSNs have little interest in dealing with information accuracy because such tasks are resource and time consuming and might contradict or limit their main goal of promoting information sharing. Conversely, there are some websites that attempt to track accuracy, but they tend to use largely manual methods to evaluate contents. More importantly, do such efforts work? Here we propose a crowd-based collaborative model for information-credibility assessment. The model strikes a balance between social reach and the motivations to create more and more content on the one hand and quality and credibility in such content on the other, with assessment occurring through user credibility, a user's own peers, and/or networks and content cited websites.

format_quoteIdentified conflicting goals in limiting fake news spread, such as resource constraints and exposure timing.format_quote

Download

Enhancing Neural Text Detector Robustness with μAttacking and RR-Training

Electronics

With advanced neural network techniques, language models can generate content that looks genuinel... more With advanced neural network techniques, language models can generate content that looks genuinely created by humans. Such advanced progress benefits society in numerous ways. However, it may also bring us threats that we have not seen before. A neural text detector is a classification model that separates machine-generated text from human-written ones. Unfortunately, a pretrained neural text detector may be vulnerable to adversarial attack, aiming to fool the detector into making wrong classification decisions. Through this work, we propose μAttacking, a mutation-based general framework that can be used to evaluate the robustness of neural text detectors systematically. Our experiments demonstrate that μAttacking identifies the detector’s flaws effectively. Inspired by the insightful information revealed by μAttacking, we also propose an RR-training strategy, a straightforward but effective method to improve the robustness of neural text detectors through finetuning. Compared with ...

Download

Models versus Datasets: Reducing Bias through Building a Comprehensive IDS Benchmark

Future Internet, 2021

This article is an open access article distributed under the terms and conditions of the Creative... more

Download

New Waves of Cyber Attacks in the Time of COVID19

Advances in Computing, Informatics, Networking and Cybersecurity, 2022

For hackers and many actors in the cyber world, crises can make opportunities. With the continuou... more For hackers and many actors in the cyber world, crises can make opportunities. With the continuous spread of Covid-19 pandemics, new waves of cyber-attacks are witnessed. Attackers are taking advantage of possibly increasing vulnerabilities due to lack of awareness of best security practices when using online resources by many new users. While the data and network security specialist do their best to protect the cyber users, and Information Technology companies strive to look for threats countermeasures, in this chapter, we are presenting, revising, and analyzing different types of cyberattacks which might exist before, but evolved at the COVID-19 era. Moreover, we are addressing the new cyberattacks that flourished during COVID-19 pandemic from different prospective including their economical and social impact. Finally, we are discussing possible countermeasures to detect and prevent these attacks. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Misconfiguration in Firewalls and Network Access Controls: Literature Review

Future Internet, 2021

Firewalls and network access controls play important roles in security control and protection. Th... more Firewalls and network access controls play important roles in security control and protection. Those firewalls may create an incorrect sense or state of protection if they are improperly configured. One of the major configuration problems in firewalls is related to misconfiguration in the access control roles added to the firewall that will control network traffic. In this paper, we evaluated recent research trends and open challenges related to firewalls and access controls in general and misconfiguration problems in particular. With the recent advances in next-generation (NG) firewalls, firewall roles can be auto-generated based on networks and threats. Nonetheless, and due to the large number of roles in any medium to large networks, roles’ misconfiguration may occur for several reasons and will impact the performance of the firewall and overall network and protection efficiency.

Download

Adversarial Attacks and Defenses for Social Network Text Processing Applications: Techniques, Challenges and Future Research Directions

The growing use of social media has led to the development of several Machine Learning (ML) and N... more The growing use of social media has led to the development of several Machine Learning (ML) and Natural Language Processing (NLP) tools to process the unprecedented amount of social media content to make actionable decisions. However, these ML and NLP algorithms have been widely shown to be vulnerable to adversarial attacks. These vulnerabilities allow adversaries to launch a diversified set of adversarial attacks on these algorithms in different applications of social media text processing. In this paper, we provide a comprehensive review of the main approaches for adversarial attacks and defenses in the context of social media applications with a particular focus on key challenges and future research directions. In detail, we cover literature on six key applications, namely (i) rumors detection, (ii) satires detection, (iii) clickbaits & spams identification, (iv) hate speech detection, (v) misinformation detection, and (vi) sentiment analysis. We then highlight the concurrent and...

Download

Analysis and Prediction of COVID-19 Timeline and Infection Rates

2020 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM), 2020

The number of new cases of infection of the Coronavirus disease, COVID-19, is alarming in many pl... more The number of new cases of infection of the Coronavirus disease, COVID-19, is alarming in many places in the world. In several world countries, including USA, the infection rates and daily cases numbers are fairly high; and there are even some spike increases in some USA states. Since the USA is experiencing the highest number of daily new cases in the world from May through July, the most important question is when we will witness an effective decline in the number of daily new cases? This paper follows a data-driven approach to induce the disease decline values from two country groups in the world where the disease declined already to less than 25% of its peak daily new cases. We apply these country groups' models to predict the decline to 25% of the US's peak. We compiled, examined, and analyzed pandemic data and statistics of two countries: g1: 42 countries, and g2: 14 countries. We utilize their data in the prediction of the decline timeline of the US. Group g2 consists of 14 countries having a similar number of cases per one million population. The majority of the models predict that the decline to 25% of the US's peak will be around the end of November to the first week of October. The results are significant and impressive as it is highly demanded to have clues and methods for the timeline prediction of this pandemic in the USA.

format_quoteQuestions arise about when the decline in daily COVID-19 cases in the USA will occur amidst rising infections.format_quote

Download

Event Detection in Twitter: A Content and Time-Based Analysis

ArXiv, 2021

The detection of events from online social networks is a recent, evolving field that attracts res... more The detection of events from online social networks is a recent, evolving field that attracts researchers from across a spectrum of disciplines and domains. Here we report a time-series analysis for predicting events. In particular, we evaluated the frequency distribution of top n-grams of terms over time, focusing on two indicators: high-frequency n-grams over both short and long periods of time. Both indicators can refer to certain aspects of events as they evolve. To evaluate the model’s accuracy in detecting events, we built and used a Twitter dataset of the mostpopular hashtags that surrounded the well-documented protests that occurred at the University of Missouri (Mizzou) in late 2015.

Download

Using Data Analytics to Filter Insincere Posts from Online Social Networks A Case Study: Quora Insincere Questions

Proceedings of the Annual Hawaii International Conference on System Sciences, 2020

The internet in general and Online Social Networks (OSNs) in particular continue to play a signif... more The internet in general and Online Social Networks (OSNs) in particular continue to play a significant role in our life where information is massively uploaded and exchanged. With such high importance and attention, abuses of such media of communication for different purposes are common. Driven by goals such as marketing and financial gains, some users use OSNs to post their misleading or insincere content. In this context, we utilized a real-world dataset posted by Quora in Kaggle.com to evaluate different mechanisms and algorithms to filter insincere and spam contents. We evaluated different preprocessing and analysis models. Moreover, we analyzed the cognitive efforts users made in writing their posts and whether that can improve the prediction accuracy. We reported the best models in terms of insincerity prediction accuracy.

format_quoteData analytics identified crucial for developing scalable models to detect misleading content across social media platforms.format_quote

Download

A prototype for a standard arabic sentiment analysis corpus

Int. Arab J. Inf. Technol., 2016

The researchers in the field of Arabic sentiment analysis (SA) need a relatively big standard Ara... more The researchers in the field of Arabic sentiment analysis (SA) need a relatively big standard Arabic sentiment analysis corpus to conduct their studies. There are a number of existing Arabic datasets; however they suffer from certain limitations such as the small number of reviews or topics they contain, the restriction to Modern Standard Arabic (MSA), or not being publicly available. Therefore, this study aims to establish a flexible and relatively big standard Arabic sentiment analysis corpus that can be considered as a pillar and cornerstone to build larger Arabic corpora. In addition to MSA, this corpus contains reviews written in the five main Arabic dialects (Egyptian, Levantine, Arabian Peninsula, Mesopotamian, and Maghrebi group). Furthermore, this corpus has other five types of reviews (English, mixed MSA & English, French, mixed MSA & Emoticons, and mixed Egyptian & Emoticons). This corpus is released for free to be used by researchers in this field, where it is characteri...

Download

Aspect-Based Sentiment Analysis of Arabic Laptop Reviews

Sentiment Analysis (SA) is one of the hottest research areas in Natural Language Processing (NLP)... more Sentiment Analysis (SA) is one of the hottest research areas in Natural Language Processing (NLP) with vast commercial as well as academic applications. One of the most interesting versions of SA is called Aspect-Based SA (ABSA). Currently, most of the researchers focus on English text. Other languages such as Arabic have received less attention. To the best of our knowledge, only few papers have addressed ABSA of Arabic reviews and they have all been applied on only three datasets. In this work, we demonstrate our efforts to build the Arabic Laptops Reviews (ALR) dataset, which focuses on laptops reviews written in Arabic. To make it easy to use, the ALR dataset is prepared according to the annotation scheme of SemEval16-Task5. The annotation scheme considers two problems: aspect category prediction and sentiment polarity label prediction. It also comes with an evaluation procedure that extracts n-grams’ features and employs a Support Vector Machine (SVM) classifier in order to all...

Download

Extended Topical Classification of Hadith Arabic Text

Applications of automatic text classification to one of many possible categories extend frequentl... more Applications of automatic text classification to one of many possible categories extend frequently in many information system and domain areas. Research on different methods of text classification evaluates the quality of this classification in terms of performance or accuracy metrics. In this paper we evaluated the automatic classification of Islam prophet sayings based on several possible categories. This study aims to evaluate the effectiveness of four well-known classification algorithms (Naive Bayes ( NB ), Bagging, Support Vector Machine ( SVM ) and LogiBoost) to classify Prophet Mohammed (Peace and blessings of Allah be upon him (PBUH)) Arabic text sayings into one of five classes (books) ((Ablutions (Wudu'), Fasting, Almsgiving (Zakat), Prayers, and Call to Prayers (Adhaan))). This study is based on (Sahih Bukhari, " صحيح البخاري ") collection that is considered the first of (The Two Authentic, " الصحيحين ") collections. Evaluation results showed that...

$Table 3: Evaluation Results of the LogiBoost classifier Table 3 shows that LogiBoost classifier yields good accuracy rates, where the True Positive accuracy rate of Call to Prayers (Adhaan) is high (0.882), while the accuracy rates for the other four classes were very low especially the accuracy rate of Prayers class (0.182). LogiBoost classifier yields an accuracy of 53.0997 %, and an error rate of 46.9003 % for the dataset used in this study. Table 3 presents the values of different measurements that are used to evaluate the effectiveness of LogiBoost classifier to classify ( Ahadith, "“:s!"/Sayings) of the dataset into the five selected classes (books). Where 7P is the number of C Ahadith, "“s4!"/Sayings) correctly classified as belonging to a class i (“true positive”), FP is the number of (’Ahadith, "“s+4s\"/Sayings) falsely classified as belonging to a class i (“false positive”) and FN is the number of (’ Ahadith, "“2lsi"/Sayings) falsely classified as not belonging to a class i (“false negative”) (Witten, Frank, & Hall, 2011).$

$Table 4: Evaluation Results of the Bagging classifier ‘able 4 shows that the effectiveness of Bagging classifier is higher than the effectiveness of ogiBoost classifier, but generally the results of Bagging classifier still lower than 50%. The esults of Bagging classifier in table 4 showed that the True Positive Accuracy rate of the ollowing four classes out of five classes are lower than 50%: Ablutions (Wudu’), Fasting, \lImsgiving (Zakat), and Prayers.), are (0.029), (0.254), (0.407), and (0.114). Table 4 has mly one exceptional high accuracy rate (0.854) for the Call to Prayers (Adhaan) class, and his led to higher overall accuracy percentage for the Bagging classifier over the accuracy of . classes.$

Download

An ontological analysis of misinformation in online social networks

The internet, Online Social Networks (OSNs) and smart phones enable users to create tremendous am... more The internet, Online Social Networks (OSNs) and smart phones enable users to create tremendous amount of information. Users who search for general or specific knowledge may not have these days problems of information scarce but misinformation. Misinformation nowadays can refer to a continuous spectrum between what can be seen as "facts" or "truth", if humans agree on the existence of such, to false information that everyone agree that it is false. In this paper, we will look at this spectrum of information/misinformation and compare between some of the major relevant concepts. While few fact-checking websites exist to evaluate news articles or some of the popular claims people exchange, nonetheless this can be seen as a little effort in the mission to tag online information with their "proper" category or label.

format_quoteThe rise of OSNs has transformed individuals into content creators, radically shifting control of information from traditional media.format_quote

Download

Evaluating network test scenarios for network simulators systems

International Journal of Distributed Sensor Networks, 2017

Networks continue to grow as industries use both wired and wireless networks. Creating experiment... more Networks continue to grow as industries use both wired and wireless networks. Creating experiments to test those networks can be very expensive if conducted on production networks; therefore, the evaluation of networks and their performance is usually conducted using emulation. This growing reliance on simulation raises the risk of correctness and validation. Today, many network simulators have widely varying focuses and are employed in different fields of research. The trustworthiness of results produced from simulation models must be investigated. The goal of this work is first to compare and assess the performance of three prominent network simulators-NS-2, NS-3, and OMNet++-by considering the following qualitative characteristics: architectural design, correctness, performance, usability, features, and trends. Second, introduce the concept of mutation testing to design the appropriate network scenarios to be used for protocol evaluation. Many works still doubt if used scenarios can suit well to claim conclusions about protocol performance and effectiveness. A large-scale simulation model was implemented using ad hoc on-demand distance vector and destination-sequenced distance vector routing protocols to compare performance, correctness, and usability. This study addresses an interesting question about the validation process: ''Are you building the right simulation model in the right environment?'' In conclusion, network simulation alone cannot determine the correctness and usefulness of the implemented protocol. Software testing approaches should be considered to validate the quality of the network model and test scenarios being used.