Abu Nayem

Work place: Department of Computer Science and Engineering, Stamford University Bangladesh

E-mail: jacknayeem@gmail.com

Website: https://orcid.org/0000-0001-9714-6353

Research Interests: Computer Architecture and Organization, Theoretical Computer Science, Data Structures


Abu Nayem was born in Noakhali, Bangladesh. He has completed his BSc in CSE at Stamford University Bangladesh in 2019. Currently, he is working on NLP and HCI. His research interest is in Networking, Data Science and Machine Learning.

Author Articles
How do Machine Learning Algorithms Effectively Classify Toxic Comments? An Empirical Analysis

By Md. Abdur Rahman Abu Nayem Mahfida Amjad Md. Saeed Siddik

DOI: https://doi.org/10.5815/ijisa.2023.04.01, Pub. Date: 8 Aug. 2023

Toxic comments on social media platforms, news portals, and online forums are impolite, insulting, or unreasonable that usually make other users leave a conversation. Due to the significant number of comments, it is impractical to moderate them manually. Therefore, online service providers use the automatic detection of toxicity using Machine Learning (ML) algorithms. However, the model's toxicity identification performance relies on the best combination of classifier and feature extraction techniques. In this empirical study, we set up a comparison environment for toxic comment classification using 15 frequently used supervised ML classifiers with the four most prominent feature extraction schemes. We considered the publicly available Jigsaw dataset on toxic comments written by human users. We tested, analyzed and compared with every pair of investigated classifiers and finally reported a conclusion. We used the accuracy and area under the ROC curve as the evaluation metrics. We revealed that Logistic Regression and AdaBoost are the best toxic comment classifiers. The average accuracy of Logistic Regression and AdaBoost is 0.895 and 0.893, respectively, where both achieved the same area under the ROC curve score (i.e., 0.828). Therefore, the primary takeaway of this study is that the Logistic Regression and Adaboost leveraging BoW, TF-IDF, or Hashing features can perform sufficiently for toxic comment classification.

[...] Read more.
Non-Functional Requirements Classification Using Machine Learning Algorithms

By Abdur Rahman Abu Nayem Saeed Siddik

DOI: https://doi.org/10.5815/ijisa.2023.03.05, Pub. Date: 8 Jun. 2023

Non-functional requirements define the quality attribute of a software application, which are necessary to identify in the early stage of software development life cycle. Researchers proposed automatic software Non-functional requirement classification using several Machine Learning (ML) algorithms with a combination of various vectorization techniques. However, using the best combination in Non-functional requirement classification still needs to be clarified. In this paper, we examined whether different combinations of feature extraction techniques and ML algorithms varied in the non-functional requirements classification performance. We also reported the best approach for classifying Non-functional requirements. We conducted the comparative analysis on a publicly available PROMISE_exp dataset containing labelled functional and Non-functional requirements. Initially, we normalized the textual requirements from the dataset; then extracted features through Bag of Words (BoW), Term Frequency and Inverse Document Frequency (TF-IDF), Hashing and Chi-Squared vectorization methods. Finally, we executed the 15 most popular ML algorithms to classify the requirements. The novelty of this work is the empirical analysis to find out the best combination of ML classifier with appropriate vectorization technique, which helps developers to detect Non-functional requirements early and take precise steps. We found that the linear support vector classifier and TF-IDF combination outperform any combinations with an F1-score of 81.5%.

[...] Read more.
Security Analysis of Government & Financial Websites of Bangladesh

By Md. Asaduzzaman Masum Md. Rishad Istiak Sachcha Abu Nayem

DOI: https://doi.org/10.5815/ijeme.2022.02.03, Pub. Date: 8 Apr. 2022

The vision 2021 of Bangladesh had to transform into a digital country, where the digital platform was a significant part of it. To make a digital platform, the Bangladesh government announced plans to build web applications in government, non-government, financial, educational and other sectors. By increasing the number of websites, the security risk is growing because of vulnerable coding practices. If those security risks are not fixed, attackers could exploit these vulnerabilities and perform various malpractices like data breaches, injected spam content, spreading viruses, malicious redirects, Denial-of-service, or even website defacements. This paper focuses on vulnerability assessment on Bangladeshi government and financial websites to show the security posture of these sites. This study scanned and analyzed four types of risk alerts High, Medium, Low and Informational using Acunetix and ZAP tools. In addition, the selected top five vulnerabilities are CJ, MC, CSRF, ID and XSS in terms of single vulnerability-type detected for targeted websites. The report has described representing the security condition of Bangladesh official websites. Also, it provided mitigation techniques for these vulnerabilities to avoid security risk, which is less discussed in this country.

[...] Read more.
Other Articles