Mohd Fazil

Dr. Mohd Fazil

orcid

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
11 April 2018 in IEEE Transactions on Information Forensics and Security
Verified
Journal Article
Closed access
MFMohd Fazil
Mohd Fazil
MAMuhammad Abulaish
Muhammad Abulaish
Twitter is one of the most popular microblogging services, which is generally used to share news and updates through short messages restricted to 280 characters. However, its open nature and large user base are frequently exploited by automated spammers, content polluters, and other ill-intended users to commit various cybercrimes, such as cyberbullying, trolling, rumor dissemination, and stalking. Accordingly, a number of approaches have been proposed by researchers to address these problems. However, most of these approaches are based on user characterization and completely disregarding mutual interactions. In this paper, we present a hybrid approach for detecting automated spammers by amalgamating community-based features with other feature categories, namely metadata-, content-, and interaction-based features. The novelty of the proposed approach lies in the characterization of users based on their interactions with their followers given that a user can evade features that are related to his/her own activities, but evading those based on the followers is difficult. Nineteen different features, including six newly defined features and two redefined features, are identified for learning three classifiers, namely, random forest, decision tree, and Bayesian network, on a real dataset that comprises benign users and spammers. The discrimination power of different feature categories is also analyzed, and interaction- and community-based features are determined to be the most effective for spam detection, whereas metadata-based features are proven to be the least effective.

12 May 2024 in Journal of Learning Analytics
Verified
Journal Article
Open access
MFMohd Fazil
Mohd Fazil
ARAngélica Rísquez
Angélica Rísquez
CHClaire Halpin
Claire Halpin
Technology-enhanced learning supported by virtual learning environments (VLEs) facilitates tutors and students.VLE platforms contain a wealth of information that can be used to mine insight regarding students’ learning behaviourand relationships between behaviour and academic performance, as well as to model data-driven decision-making.This study introduces a system that we termed ASIST: a novel Attention-aware convolutional Stacked BiLSTM net-work for student representation learning to predict their performance. ASIST exploits student academic registry, VLEclick stream, and midterm continuous assessment information for their behaviour representation learning. ASISTjointly learns the student representation using five behaviour vectors. It processes the four sequential behaviourvectors using a separate stacked bidirectional long short term memory (LSTM) network. A deep convolutional neuralnetwork models the diurnal weekly interaction behaviour. It also employs the attention mechanism to assign weightto features based on their importance. Next, five encoded feature vectors are concatenated with the assessmentinformation, and, finally, a softmax layer predicts the high-performer (H), moderate-performer (M), and at-risk (F)categories of students. We evaluate ASIST over three datasets from an Irish university, considering five evaluationmetrics. ASIST achieves an area under the curve (AUC) score of 0.86 to 0.90 over the three datasets. It outperformsthree baseline deep learning models and four traditional classification models. We also found that the attentionmechanism has a slight impact on ASIST’s performance. The ablation analysis reveals that weekly event count hasthe greatest impact on ASIST, whereas diurnal weekly interaction has the least impact. The early prediction usingthe first seven weeks of data achieves an AUC of 0.83 up to 0.89 over the three datasets. In yearly analysis, ASISTperforms best over the 2018/19 dataset and worst over the 2020/21 dataset.

1 January 2017
Verified
Conference Paper
Closed access
MFMohd Fazil
Mohd Fazil
MAMuhammad Abulaish
Muhammad Abulaish
Twitter, a popular microblogging platform, facilitates users to express views and thoughts on any topic of discussion using short messaging texts limited to 140 characters. Due to its open and real-time information sharing and dissemination nature, it is abused by socialbots for political astroturfing, advertising, spamming, and other illicit activities. To this end, we injected an army of 98 socialbots associated to top six Twitter using countries to study socialbots' infiltration behaviour. In this paper, we present a statistical insight derived through the analysis of the captured data by our socialbots. Socialbots' profile features, such as age, gender, etc. and their behavioural impact on infiltration performance are studied and presented, wherein a user's following activity to a socialbot is considered as an infiltration. Experimental results and subsequent statistical analyses show that socialbots' profiles belonging to India were the successful in duping highest number of users, whereas Indonesian socialbots were least infiltrative. Moreover, among various Twitter activities, following is found to be the most effective activity for infiltrating a user. Among the intruded users, trace of the presence of botnets, spammers, and other malicious users have also been observed and presented in this paper.

23 August 2017
Verified
Conference Paper
Open access
MFMohd Fazil
Mohd Fazil
MAMuhammad Abulaish
Muhammad Abulaish
Online social networks are facing serious threats due to presence of human-behaviour imitating malicious bots (aka socialbots) that are successful mainly due to existence of their duped followers. In this paper, we propose an approach to categorize Twitter users into three groups - active, reactive, and inactive targets, based on their interaction behaviour with socialbots. Active users are those who themselves follow socialbots without being followed by them, reactive users respond to the following socialbots by following them back, whereas inactive users do not show any interest against the following requests from anonymous socialbots. The proposed approach is modelled as both binary and ternary classification problem, wherein users' profile is generated using static and dynamic components representing their identical and behavioural aspects. Three different classification techniques viz Naive Bayes, Reduced Error Pruned Decision Tree, and Random Forest are used over a dataset of 749 users collected through live experiment, and a thorough analyses of the identified users categories is presented, wherein it is found that active and reactive users keep on frequently updating their tweets containing advertising related contents. Finally, feature ranking algorithms are used to rank identified features to analyse their discriminative power, and it is found that following rate and follower rate are the most dominating features.

1 January 2019
Verified
Conference Paper
Closed access
MAMuhmmad Abulaish
Muhmmad Abulaish
SSSielvie Sharma
Sielvie Sharma
MFMohd Fazil
Mohd Fazil
The popularity of the microblogging sites like Twitter is increasing exponentially in which users are allowed to post short messages (aka tweets) using a maximum of 280 characters, mainly for news sharing and events updates. Besides, textual data, Twitter data also contains multi-dimensional connections among the users if they follow each other or have common followers/followees. Similarly, multi-dimensional connections exist among the tweets if they contain common hashtags, mentions, etc. In recent years, Word2Vec is being extensively used to analyze textual data, and it has shown promising results in many domains. In this paper, we propose a multi-attributed graph-based approach for text data modeling and event detection in Twitter. To this end, we generate a multi-attributed social graph (MASG), in which nodes, representing tweets, are labelled with numeric vectors obtained through Word2Vec model, and edges represent the structural relationships among the tweets and they can also be labeled with numeric vectors. Thereafter, MASG is converted into a similarity graph using a distance function, and Markov clustering algorithm is applied over the similarity graph to identify different clusters, where each cluster corresponds to a particular event. The proposed approach is evaluated over real-world Twitter datasets using standard evaluation metrics including TPR and FPR. It is also compared with some baseline methods and performs significantly better.

29 October 2018 in IEEE Access
Verified
Journal Article
Open access
MAMuhammad Abulaish
Muhammad Abulaish
MFMohd Fazil
Mohd Fazil
In last two decades, online social networks have grown vertically as well as horizontally. Due to various users activities in these networks, huge amount of data, mainly textual, is being generated that can be analyzed at different levels of granularity for various purposes, including behavior analysis, sentiment analysis, and predictive modeling. In this paper, we propose a word embedding-based approach to analyze users-centric tweets to observe their behavior evolution in terms of the topics discussed by them over a period of time. We also present a word embedding-based proximity measure to monitor temporal transitions between the topics using five topic evolution events – emergence, persistence, convergence, divergence, and extinction. The proximity between a pair of topics is defined as a function of the content and contextual similarity between their word distributions, wherein contextual similarity is calculated using word embedding. The proposed approach is evaluated over three Twitter datasets in line with the existing state-of-the-art approaches in literature and the experimental results are encouraging.