Browsing by Author "Akhtaruzzaman, M."

Now showing 1 - 9 of 9

A Neural Network Based Software Defect Prediction Approach Using SMOTE and Noise Filtering-CLNI
(Research and Development Wing, MIST, 2025-12-30) Ashfaque, Ahmmed Bin; Sattar, Abdus; Jahan, Hosney; Akhtaruzzaman, M.; Nur, Fernaz Narin
Software defects can cause significant loss and system failures in software development life cycle. Software Defect Prediction (SDP) is a vital step for ensuring the quality of software. Till now, a number of machine learning models have been proposed to predict potential defects and make the software more reliable. However, SDP models suffer from the problem of imbalanced dataset, resulting in poor prediction accuracy. To mitigate this, issue several data balancing techniques, i.e., over sampling, under sampling etc. have been proposed to balance the dataset. In some cases, the data balancing methods may further introduce noisy and mislabeled samples in the dataset. To deal with these issues, in this paper, we propose a neural network based approach that combines the oversampling technique Synthetic Minority Oversampling Technique (SMOTE) with the noise filtering technique Class Level Noise Identification (CLNI). Here, we applied three different CLNI methods which are Edited Nearest Neighbor (ENN), Repeated ENN (RENN) and All-KNN. Our aim is to make the dataset clean, balanced and efficient by combining SMOTE with CLNI. In addition, we applied a number of feature selection methods to identify the most important features, further contributing towards achieving better prediction accuracy. To evaluate the effectiveness of the proposed model, we conduct experiments on several benchmark datasets (MC1, PC1, PC2, PC3 and PC4) obtained from NASA MDP and (ML, LC and JDT) AEEEM repository. The experimental results have been evaluated and compared in terms of accuracy, precision, recall and AUC-ROC curve. The experimental results demonstrated that our proposed approach has achieved up to 98% accuracy and outperformed state-of- the-art approaches.
Analyzing the Performance of Deep Learning Models for Detecting Hate Speech on Social Media Platforms
(Research and Development Wing, MIST, 2024-12) Islam Arif, Md Ariful; Rahman, Md. Mahbubur; Rabiul Alam, Md. Golam; Akhtaruzzaman, M.
Currently social media and online platforms have become a major source of cyberbullying and hate speech. It is currently affecting people and communities in harmful ways. Hate speech on social media is rising in Bangladesh and it is creating a need for effective tools to prevent and detect these incidents. This study introduces a deep learning model to mitigate this issue of identifying hate speech in text using three types of word embedding methods: Word2Vec, FastText, and BERT. The text data was labeled to mark hate speech and non-hate speech content. After that, these texts are preprocessed by removing punctuation and symbols to help improve model accuracy. Five deep learning models Bi-GRU-LSTM-CNN, Bi-LSTM, CNN, LSTM, and XGBoost were trained to classify the text as hate speech or non-hate speech. The study found that the LSTM model accomplished the highest accuracy at 95.66% with the Word2Vec embedding method, while CNN reached 87.70% with FastText embeddings. Word2Vec is effective for capturing word meanings in general text classification. FastText works well with rare words and languages that have complex word forms. These findings help advance effective hate speech detection techniques. It could promote more respectful and inclusive interactions on social media. This proposed deep-learning model can help stop cyberbullying and hate speech on social media.
Decentralized LRM System Architecture with Biometric Authentication and Digital Certificate Verification through Blockchain Technology
(Research and Development Wing, MIST, 2025-12-30) Mosharrof, Shakil; Nizami, Farhan Nasif; Mohtasim, Mahdi; Adib, Mahdi; Akhtaruzzaman, M.; Islam, Md Shofiqul; Rahman, Muhammad Towfiqur; Jahan, Hosney
Managing land records is a fundamental duty of a government, ensuring the accuracy, consistency, integrity of ownership, and reliable transaction of data. Conventional paper-based or centralized digital technology-based land record systems are unable to hold the system trust, efficiency, and consistency, thus mostly demonstrate errors, fraud, and corruptions. On the other hand, blockchain technology presents a transformative solution. It offers transparent, tamper-proof, secure, and reliable approach for Land Record Management (LRM) system. In this study a blockchaindriven LRM system architecture with distributed ledger technology is presented. The proposed strategy enhances security and trust by ensuring transparency, acceptance, and accountability. The study also designs the smart-contracts algorithm in detail that facilitates land registration, ownership transfers, verification, and streamlining processes. The proposed architecture ensures automated functionalities with little human intervention, uplifting the system security. Moreover, the proposed blockchain architecture integrates finger-print biometric authentication that boosts the system strength in terms of security through identity verification. This mitigates the risks of errors, fraud, illegal modification, and unauthorized access. This article outlines a blockchain-based LRM framework and verified through implementation and testing, reflecting the potentiality of viable adoption of this advanced technology.
Food Composition Tables and Database for Bangladesh with Special Reference to Selected Ethnic Foods
(University of Dhaka, 2012) Islam, Sheikh Nazrul; Khan, Md. Nazrul Islam; Akhtaruzzaman, M.
Intelligent Head-bot, towards the Development of an AI Based Cognitive Platform
(Research and Development Wing, MIST, 2023-12) Baki, Ramisha Fariha; Akhtaruzzaman, M.; Refat, Tahsin Ahmed; Rahman, Mouneeta; Razzak, Md Abdur; Majumder, Md Mahfuzul Karim; Islam, Md Adnanul; Ferdaus, Md Meftahul; Rahman, Muhammad Towfiqur; Naveed, Quadri Noorulhasan
A cognitive humanoid head is an AI enabledhead-bot platformthat resembleshuman's cognitive abilities, such as perception, thinking, learning, and decision-making. The platformis able tointeract with human throughnatural language processingand recognizeindividuals, thusallowing seamless communication between two parties. No such cognitive platform has been introduced inBangladesh,thuscreatinganopenfieldtocontribute to the field of Machine Intelligence.Thisstudy aims to develop an AIbased humanoid head (head-bot) capable of imitating arange of expressions, recognizing individuals, and interacts with visitorsthrough general conversation.The head-botskeleton is developed using a number of hexagonal blocks of PVC sheet to mimic a human-head-like structure where LCD, camera, microphones, and speaker are mounted. Twoseparate Machine Learningmodels are designed for face detection andrecognitions, and voice enabledchat-bot implementation. The head-botplatform incorporates 2-DoF neck movements for various head gestures and face tracking.The Artificial Neural Network models are tested with accuracy of , and , for face detection andrecognitions, and speech recognitionsand response generation, respectively. According to the overall results and system performances, it seems that the proposed system has a number of good potentialsfor real life applicationssuch as entertainment, guidance, conversations, interactive receptionists, personal companion, medical assistance, and so on.
Link Budget Analysis in Designing a Web-application Tool for Military X-Band Satellite Communication
(R&D Wing, MIST, 2020-06) Akhtaruzzaman, M.; Sadakatul Bari, S. M.; Hossain, Syed Akhter; Rahman, Md. Mahbubur
In satellite communication, Link Budget analysis is the most important part todeterminegainsand lossesof signalsfrom the transmitter to the receiver. Most importantly, it investigates system performance and optimum power which must be received at the receiverchannel. In some cases, this information could be generated, saved for past data analysis, and share with peer users which are not found in existing web tools. Thus, it is obvious to design a newLink Budget calculator with users, database, and data retrieval support. This work focuses on designingaLink Budget web tool for X-band satellite communication through literaturestudy and comparative analysis. The X-Band calculatoris designed based on HTML, PHP, Javascript, and MySQLby ensuring several security issues, andcan be accessedthrough mobile devices.This paperalso focuseson the necessary equations of Link Budget forUplink (푇푥);Satellite;Downlink (푅푥); Azimuth, Elevation,Distance analysis; andRain attenuation. Though, comparative assessments among various web tools show some fluctuations, overall outputs show satisfactory results with small% of Errors (PoE) ensuringreliability and viability of the proposed X-Band tool for practical use.
Modeling and Control Simulation of a Robotic Chair-Arm: Protection against COVID-19 in Rehabilitation Exercise
(R&D Wing, MIST, 2020-12-16) Akhtaruzzaman, M.; Shafie, Amir A.; Khan, Md Raisuddin; Rahman, Md Mozasser
In the field of rehabilitation, lower-limbs therapeutic exercise has become a challenging job for medical professionals in COVID-19 pandemic. Providing manual therapy to lower limbs is not an easy task and, in most cases, it involves multiple persons. Moreover, it is a monotonous job, and the service providers need to be in close contact with the patient thereby creating the risk of infection. In this circumstance, robot-assisted rehabilitation exercise for lower limbs offers a risk-free solution. This paper presents dynamic modeling and control simulation of One Degree of Freedom robotic chair-arm (robotic arm attached with a special chair). The control structure is designed with two compensators for position and velocity control. The simulation results show that the proposed system has a good potential in providing automatic rehabilitation therapy for lower limbs, especially for knee joint range of motion exercise. The results also indicate faster responses with settling time less than 0.04 second and steady-state error below 0.05. The findings show that a robotic chair arm can be used for providing automatic therapy to patients in situations like COVID-19 pandemic.
Nutrition, Health and Demographic Survey of Bangladesh-2011
(Institute of Nutrition and Food Science, University of Dhaka, 2013-12) Akhtaruzzaman, M.; Nazrul Islam Khan, Md.; Aktar, Fahmida; Islam, Sheikh Nazrul
Key Findings The NHDSBD-2011 is the fifth national survey addressing nutrition, health and demographic issues of the mass people of Bangladesh, It is the largest household survey comprising -7000 households with 31066 populations. This survey provides updated information on nutrition, health and demographic profile, and social progress including socioeconomic condition, food security, sanitation and hygiene, child and maternal health and nutrition, family planning, women empowerment, domestic violence, AIDS/STDs/STIs/TB/NCDs related knowledge, attitude and prevalence.
System Usability and Design Evaluation of AI Chatbots: A Comparative Analysis of ChatGPT, Google Bard, and Bing Chat
(Research and Development Wing, MIST, 2025-06) Mustafina, Sumaiya Nuha; Khan, Nusrat Kaniz; Islam, Muhammad Nazrul; Siddiqua Nusrat, Fatema; Akhtaruzzaman, M.
Artificial intelligence (AI) has brought significant advancements in technology while the chatbots like ChatGPT, Google Bard, and Bing Chat are some of its remarkable innovations. These chatbots are helping users with diverse backgrounds by generating ideas, providing resources, and overall knowledge management. We acknowledge that these chatbots are still in their experimental stages of use. Evaluating the usability and user experience of chatbots becomes crucial to make them more usable, accessible, and intuitive to end users around the globe. Thus, the objectives of this research are to make a comparative usability analysis of AI-generated chatbots: Google Bard, ChatGPT, and BingChat. To achieve these goals, firstly, the System Usability Score (SUS) through questionnaire surveys and secondly, Heuristic Evaluation (HE) through expert observation were used. Through HE, we investigated characteristics of design, user engagement, and some other specific usability lacking along with a severity score that suggests both urgent and gradual usability improvement action. As an outcome, this study found that the SUS evaluation provided a comprehensive view of user satisfaction. Google Bard and Bing Chat received lower SUSscores, while ChatGPT demonstrated comparatively better usability, with a SUS score above 70. Again, a comparative usability analysis of AI-generated chatbots (ChatGPT, Google Bardand Bing Chat) reveals that, while all these applications suffer from a notable number of usability problems, ChatGPT demonstrates better usability performance compared to Google Bard and BingChat.