All of the articles in this proceedings volume have been presented at the [International Conference on Intelligent Data Analysis and Applications (IDAA 2025)] during [12-13 December 2025] in [Daffodil International University, Dhaka-1216, Bangladesh]. These articles have been peer reviewed by the members...

Deep learning has recently garnered significant attention for developing fast, automated, and accurate image classification and identification systems. This study focuses on enhancing and evaluating state-of-the-art deep convolutional neural network (CNN) architectures for imaging-based cholangiocarcinoma...

Early diagnosis of Alzheimer’s disease (AD) significantly improves patient outcomes, yet most studies achieving greater than 95% accuracy rely on expensive MRI/PET imaging or deep learning models that are infeasible in low-resource settings. This work demonstrates that simple, interpretable machine learning...

Monkeypox detection and prevention depend on proper and timely diagnosis of the disease. Manual clinical examination is risky for healthcare staff, time-consuming, and costly. Hence, computer-aided diagnosis of monkeypox is highly valuable. In this work, multiple deep learning algorithms are employed...

Chest X-ray imaging is utilized significantly in the diagnosis of respiratory disorders like COVID-19, viral pneumonia, and lung opacities. Deep learning has evolved computerized classification systems that are able to assist radiologists in making more accurate and rapid diagnoses. In this paper, we...

Lung cancer remains the leading cause of cancer-related mortality around the world, requiring headways in early discovery methods to progress understanding results. This study explores the efficacy of deep learning models, particularly InceptionV3, VGG16, and MobileNetV2, within the detection and classification...

Chest X-ray imaging remains a frontline tool for diagnosing thoracic diseases, yet manual reading is labor-intensive and susceptible to inter-reader variability. This work proposes CXR-Next, an explainable deep learning framework built upon a ConvNeXt-Base backbone to perform six-class classification—Normal,...

Medical imaging is essential in the identification of brain tumors early and accurately in order to plan diagnosis and treatment. The latest progress in Vision Transformers (ViTs) has shown good promise in the medical image classification tasks. Nonetheless, typical ViT models typically need big data...

Accurately segmenting brain tumors from MRI scans plays a critical role in reliable diagnosis, treatment planning, and monitoring of patients. However, several key limitations persist in state-of-the-art U-Net-based approaches, such as loss of fine-grained features, poor generalization across heterogeneous...

Stroke is a leading cause of mortality and permanent disability worldwide. It is a serious neurological emergency. Timely intervention and better outcomes depend on accurate identification of its two major types, ischemic and hemorrhagic stroke. Manual interpretation of brain computed tomography (CT)...

Cerebral stroke is a leading global cause of death and long-term disability. Early, accurate identification of stroke presence and subtype (ischemic vs hemorrhagic) from brain imaging is essential because therapies differ widely (e.g., thrombolysis for ischemic vs surgery in hemorrhagic). Nevertheless,...

Early diagnosis of dengue fever is critical to patient care and outbreak control, yet reliable clinical indicators are often elusive. Machine learning (ML) offers promise by learning complex patterns in clinical and hematological data. In this study, we first review recent ML-based dengue diagnosis models...

User trust is the basis of chatbots for customer-service, especially where disclosure of personal information is in question. We frame trust prediction as supervised binary classification on questionnaire data (N=122) with questionnaire items (demographics, frequency of use, overall and recent satisfaction,...

Endometrial carcinoma is one of the most prevalent gynecological cancer in the world, recurrence of which impacts a lot on the survival of the patient. The conventional clinical markers are not usually able to render the interaction between the genomic and histopathological characteristics. In this research,...

One of the health issues that has attracted much attention in the world is high blood pressure (BP), which requires consistent blood pressure measurements. The traditional cuff devices, however dependable, cannot be used regularly and are inconvenient; as a result, Photoplethysmography (PPG) cuffless...

This research looks at the use of machine learning to predict the risk levels of antenatal complications, presenting a novel approach to improving proactive antenatal care. In this study, we have collected dataset of 800 entries with eight key attributes such as Age, Weight, BMI, systolic BP, diastolic...

The staging of chronic kidney disease (CKD) in low resource settings requires the right balance between diagnostic accuracy and computational efficiency. In this study, we propose CKDX-Net, a novel knowledge distillation pipeline transferring the patterns learned from a high-capacity XGBoost ensemble...

Cervical cancer is a leading cause of morbidity and mortality among women globally, particularly in developing countries with limited screening infrastructure. Conventional single classifier ML techniques struggle with learning complex feature interactions and extreme class imbalance in medical datasets,...

This paper presents a comparative study of traditional and generative data augmentation techniques for cervical cancer detection using the UCI Cervical Cancer Risk Factors dataset. Conventional oversampling methods, namely SMOTE and ADASYN, are evaluated alongside advanced generative approaches, including...

Cardiovascular disease (CVD) continues to be the leading cause of mortality globally. There is a need for accurate and clinically interpretable predictive systems for CVD. In this paper, we propose a multi-input Long Short-Term Memory (LSTM) model with an attention mechanism for predicting CVD, enhanced...

Liver cirrhosis is typically described as the end stage of chronic hepatic disease, in which progressive and irreversible deposition of fibrotic tissue leads to the progressive hepatic dysfunction. Early diagnosis of cirrhosis at an early stage gives a significant benefit, as it allows timely therapeutic...

Gestational diabetes mellitus is linked to adverse maternal and neonatal outcomes, which can be improved through early prediction. This current study evaluates the predictive performance of several machine learning algorithms including Logistic Regression, Decision Tree, Random Forest, Light Gradient...

This study investigates the application of machine learning (ML) algorithms in predicting depression severity among university students in Bangladesh using the Patient Health Questionnaire-9 (PHQ-9) dataset. A total of 577 students participated in the study, with data collected via an online survey that...

Sleep disorders become a public health issue in view of the association with numerous detrimental physical, cognitive, and emotional outcomes. The problem has been focused on toward predicting and classifying medical conditions related to late-night sleeping habits through supervised machine learning...

COVID-19 pandemic has left significant physical and mental health consequences with anxiety disorders being particularly prevalent among recovered patients. This study presents a machine learning framework to predict post-COVID-19 anxiety using data from 1,000 recovered individuals in Bangladesh. Several...

Mental health is one of the most important part of our overall well-being. Anxiety, depression, and emotional distress are becoming more common issues day-by-day for all age groups. People share their daily thoughts, emotions, and experiences on social media platforms such as Facebook, Twitter, and Instagram,...

There is an increasing number of depressed university students in low- and middle-income communities, yet there are few large-scale, validated tools for screening for depression. Using a structured questionnaire, we followed an inquiry protocol that included questions on everyday habits, bedtime, screen...

Detecting mental health expressions in Bangla social media text remains a critical challenge, particularly in a rapidly digitalizing society where users increasingly express emotions and psychological distress online. We used the B-MHD (Bangla Mental Health Disorder Text) dataset, a manually annotated...

Mental health problems, such as depression, anxiety and stress are becoming increasingly common among university students and private university students in Bangladesh become more vulnerable to these disorders due to many other pressures (i.e., high tuition fees, rigid academic rules and social norms)....

Mental health issues among college students have become a major worldwide issue, especially impacting young adults going through crucial transitional periods. This crisis particularly affects Bangladesh, a developing South Asian country. Students there experience high rates of anxiety, depression, and...

News classification is an important task to perform in NLP, more so when dealing with low-resource languages such as Bangla. However, Bangla comes with its own set of challenges like different morphology, complex syntax, and a very acute shortage of large annotated corpora. In this work, Long Short-Term...

The rapid worldwide spread of political discussions in Bangla calls for the automation of detection of suspicious or harmful content. This paper considers suspicious political comments in Bangla classification with both sequential-based and transformer-based deep learning names. The dataset is the Suspicious...

Posing as a diglossic and morphologically rich language, Bangla contains two major types of registers: Sadhu Bhasha, the classical type, and Cholit Bhasha, the colloquial. Identification of the registers can be beneficial for downstream applications involving NLP such as translation, OCR, and speech...

The increasing usage of Banglish-a code-mixed variety of Bangla written in Roman script-presents significant challenges for NLP. This paper presents the first cross-script framework for identifying Bengali’s two main language registers, Shadhu and Cholito, across both Bangla and Banglish text. A balanced...

A morphologically sophisticated and diglossic Bangla is a difficult language for Natural Language Processing (NLP), particularly for security tools such as smishing (SMS-based phishing) detection. This paper proposes a Bidirectional Long Short-Term Memory (BiLSTM)-based model to identify Bangla SMS as...

Class imbalance is an endemic and problematic issue in the Natural Language Processing (NLP) field, especially when Natural Language Processing is applied to low-resource languages (LRLs), in which annotated corpora often do not reflect the biased distributions of real-world hate speech. In an attempt...

With the tremendous increase of digital content in Bangla, the demand for automated document classification systems has been increasing. The complex morphology of the language, the absence of sufficient labeled documents, and the different writing styles make Bangla document classification hard to achieve....

A major challenge in Bengali Natural Language Processing (NLP) is the availability of annotated data. This research introduces an authentic, publicly available dataset collected from social networks in compliance with platform terms of service. The dataset includes 10 topic and 4 sentiment classes. The...

The rapid growth of the internet and e-commerce in Bangladesh has generated a vast repository of customer reviews in Bengali across domains such as foods, fashions and electronics. It is important to understand what users say in these multilingual user opinions, which leads to a better understanding...

In Bangladesh, motorcycle accidents are a major cause of mortality; most deaths are ascribed to riders not wearing helmets. This study intends to use a hybrid deep learning model to construct an effective, real-time helmet detection system in response to this rising issue. The system is intended to recognize...

Retinal diseases can be considered one of the leading causes of vision loss on the global scene, and OCT imaging is crucial in the timely diagnosis of the disease. In this paper, a hybrid model of deep learning, which combines Convolutional Neural Networks (CNNs), Vision Transformers (ViTs), Feature...

In both the real and virtual worlds, security is still a major concern. In order to improve access control in online meeting environments, this study suggests an effective virtual security solution that integrates face detection and recognition. The system uses the OpenCV module for face identification...

The timely clinical intervention of ischemic stroke in brain CT scans requires the early and accurate identification of its presence in the brain but this is not easy as the imaging characteristics are subtle. This paper demonstrates a well-validated deep learning model with architecture-based DenseNet121...

In recent years, various industries have been experiencing a notable increase in the use of machine learning for object recognition, resulting in the development of different methods and technologies. Thermal detection technology is one such aspect that uses thermal images captured from objects to determine...

The rapid increase of vehicles and the growth of smart transportation systems have created a strong demand for accurate and efficient Automated License Plate Recognition (ALPR) systems. This research presents an end-to-end ALPR framework optimized for Bangladeshi license plates, addressing challenges...

Road safety in Bangladesh remains a critical challenge with overspeeding and limited enforcement contributing to thousands of accidents each year. Most of the existing ANPR research focuses on plate detection alone, with none of them linking vehicle identity with driving behavior. In this paper, we bridge...

Deep vision models can perform very well on clean test sets but still break down when inputs are corrupted by noise, blur, or occlusion. We introduce Dynamic Region-Aware Gradient Suppression (DRAGS), a lightweight training-time mechanism that suppresses gradients from spatial regions detected as noisy...

Handwritten chemical structures are commonly used in chemistry education and research, but recognizing these drawings automatically remains a challenging task. Variations in writing styles, line thickness, and incomplete patterns make accurate classification difficult for computer systems. This study...

Traffic Sign Recognition (TSR) is an important part of the Advanced Driver Assistance Systems (ADAS) to guarantee intelligent vehicle safety. Correct interpretation of traffic signs can thus make humans response less prone to mistake, preventing accidents and in general increase the safety of traffic....

The application of intelligent image analysis in agriculture has improved early detection of plant diseases, helping prevent yield losses and limit infection spread. Jute, an important cash crop in Bangladesh, serves as an environmentally friendly raw material but is highly susceptible to leaf diseases...

This study proposes an overall deep learning architecture for multi-class classification of plant diseases from high-resolution leaf imagery, with a particular interest in investigating the behavior of ResNet-50 and a hybrid ResNet + Vision Transformer (ViT) design. A specially gathered image database...

Finding rice diseases is essential to maintaining the best possible crop health and reducing yield loss. By combining two different rice disease datasets and utilising transfer learning techniques, this study suggests a novel method for rice disease classification. Using the combined dataset, we optimised...

Birds are vital indicators of ecosystem health, yet numerous species in Bangladesh are threatened by habitat loss and environmental change. Existing bird classification models often employ black-box architectures lacking interpretability and dataset specificity. This paper introduces BanglaBirds-AttnNet,...

Accurate fruit ripeness detection plays a vital role in ensuring food quality, minimizing waste, and supporting automation in agriculture. This research investigates deep learning techniques for classifying the ripeness of three fruits—apples, bananas, and oranges—into fresh, unripe, and rotten categories....

Medicinal plants are a significant source of healthcare in both traditional and modern medicine, and their identification remains a significant challenge due to morphological similarities between foliage and the lack of comprehensive databases. Traditional classification methods, such as Support Vector...

Jute (Corchorus spp.) is one of the most important fibre crops for Bangladesh and many tropical countries. However, its production often suffers from leaf diseases such as insect holes, yellowing, Cercospora leaf spot, and phosphorus deficiency. Farmers usually identify these problems by visual inspection,...

Precise identification of medicinal plants is relevant to pharmacological studies and proper use of species, but most currently used image-based methods are tested on small-scale data and do not provide much information on the extrapolation of models. The paper examines how deep learning can be used...

Bridging the gap between physical farm assistance and expert agronomic advice remains an unsolved challenge. We present Amigo-Agri, an integrated platform combining a low-cost, human-following mobile robot with a retrieval-augmented voice assistant. The system splits responsibilities: a smartphone handles...

Karanda (Carissa carandas), a significant tropical fruit crop of South Asia, faces remarkable yield and quality losses due to foliar fungal and bacterial diseases. Traditional methods of disease detection are slow, subjective, and inaccurate, resulting in late interventions and significant agricultural...

Carambola (Averrhoa carambola) is one of the most important tropical fruits in terms of economy and nutrition. Production and quality of fruit crop can be severely affected by numerous foliar and fruit diseases. In this research, we propose a complete pipeline establishing a real-world multi-organ dataset...

The rapid growth of Mobile Financial Services (MFS) has increased concerns about fraud, highlighting the need for language-specific detection systems. While machine learning has advanced digital security, most solutions focus on English, leaving major languages like Bangla underserved. This study introduces...

Identifying malicious PDF files is crucial for cybersecurity since attackers are increasingly using the flexible structure and embedded content of PDFs to circumvent signature-based defenses. This work formulates a binary classification task based on interpretable machine learning on static structural...

Static zero-trust policies frequently fall short of offering adequate defense against changing threats in cloud-native microservices’ highly dynamic and distributed environments. An AI-powered zero-trust security framework that incorporates real-time anomaly detection straight into the policy-enforcement...

This study introduces a dependable method for detecting deepfake audio by combining multiple deep learning models into a single, unified system. The approach integrates two ResNet models and one CNN model, using a soft-voting strategy to merge their predictions and achieve higher overall accuracy and...

In recent years, the rapid growth of cyber threats has emphasized the importance of accurate network intrusion detection systems (NIDS). While many machine learning and deep learning models have shown promise in identifying various types of malicious traffic with the accurate classification of BENIGN...

The increasing demand for personality identification based on hand-writing processing in fields like resource management, criminal investigations, and mental health diagnostics has led to a flurry of research and experimenta-tion in this area these days. Implicit information includes characteristics...

This study presents a comprehensive spatio-temporal analysis and forecasting of crime trends in Bangladesh using monthly police data from January 2019 to June 2024. This research analyzes five crime types: theft, burglary, robbery, narcotics-related offences, and genderbased violence (GBV) across ten...

Theft and loss of personal belongings continue to be a grievous problem in Bangladesh, more so in heavily populated cities like Dhaka and Chittagong. Most of the tracking solutions are either expensive, fragmented, or poorly suited to the local infrastructure and user behaviors of this region. This paper...

As we are moving to online financial applications, it becomes necessary to have fraud detection and authentication tools. In general, the traditional methods are unable to fight against growing advanced fraud attacks and lead to greater losses in the economic industry. To the best of our knowledge, this...

Ever since artificial intelligence has gained popularity, it is being used and applied to many day-to-day things, which was quite unimaginable previously. Currently, AI tools are used to write many texts and documents. Even for preparing academic as well as business reports such platforms are being utilized....

The trend of spreading false news or rumors has kept pace with the growing utilization of social media. Hatred and fear spread through rumors, which are extremely harmful to society. As social media continues to grow, rumor detection has become an increasingly important research area. In this paper,...

Artificial Intelligence plays a vital role in generating Text that is used in our daily lives. However, the rapidly increasing usefulness of AI technology is a concern, according to need. This includes biases in the response of the generative AI. There are lots of generative AI that contribute to generating...

This paper presents a multimodal deep learning framework for humor classification in memes, leveraging both textual and visual information to improve sentiment understanding in internet content. The study explores unimodal and multimodal configurations by integrating image-based CNN architectures (MobileNetV2,...

The proliferation of AI-generated content necessitates robust detection systems across academia, journalism, and digital media. This study presents a comprehensive feature engineering framework combining 2,537 linguistic, stylometric, and semantic features for automated discrimination of AI-generated...

Generative artificial intelligence (GA) has the potential to revolutionize several industries, including the arts, entertainment, and content creation, by facilitating data synthesis and improving creativity through the use of techniques such as variational autoencoders (VAEs) and generative adversarial...

The increase in research studies focused on Large Language Models (LLMs) and Artificial Intelligence in Education (AIED) has increased the difficulty of discovering emerging themes and research trajectories. This research introduces SciClusterNet, an unsupervised methodology that combines SciBERT embeddings,...

This work introduces a Retrieval-Augmented Generation (RAG)-based scholarly assistant, for automated reading of papers, which benchmarks several open-source LLMs. The developed system uses a pipeline of document processing, citation and structural analysis, and LLM-based question-answering to produce...

Diabetes is widely recognized as a “silent killer” because of its asymptomatic onset behavior and serious long-term health consequences. Continuous high blood glucose level as an indication of the prediabetic phase. Hence, accurate and consistent glucose level monitoring is imperative for living a healthy...

Electroencephalography (EEG) signals contain rich temporalspectral structure but are difficult to model due to noise, subject variability, and multi-scale dynamics. Lightweight deep learning models have shown promise, yet many either rely solely on local convolutions or require heavy recurrent modules....

Modern battlefield environments demand intelligent systems for real-time soldier monitoring and rapid casualty response. This paper presents Intelli-Helmet, an integrated IoT and TinyML-based wearable system addressing critical gaps in military safety infrastructure. We introduce three key algorithmic...

This study introduces a unified AIoT-enabled smart aquaculture system that integrates real-time water-quality monitoring, dissolved-oxygen (DO) prediction, and fish-disease detection within a single, low-cost, field-deployable framework. The system employs Arduino sensors to measure temperature, pH,...

Fire accidents are serious life, property, and environmental hazards, and fire-fighting operations tend to put people in risky situations, particularly in hot or confined spots such as factories and narrow corridors of buildings. To address these issues, this paper documents the design and development...

To alleviate traffic congestion in the fast-urbanizing cities such as Dhaka, Bangladesh, real time prediction of short-term traffic flow should be accurately predicted. Although deep learning models like LSTM networks are great at the ability to capture the time dynamics, they become weak when exposed...

Air pollution is a growing threat to the environment in many developing countries that impacts millions of lives. Rapid urbanization and industrial activity have led to a worse impact of air pollution in Bangladesh. Accurate monitoring and forecasting are essential for public safety. This research adopts...

Emotion recognition from speech is crucial for allowing machines to comprehend and react to human emotions, which makes it extremely important for use cases like virtual assistants, healthcare diagnosis, and customer care automation. In this paper, we propose an effective and scalable emotion recognition...

The increasing demand for personality identification based on handwriting processing in fields like resource management, criminal investigations, and mental health diagnostics has led to a flurry of research and experimentation in this area these days. Implicit information includes characteristics including...

Sign language is the primary means of communication for people who are deaf. Despite the availability of enough research work for English Sign Language, research work on Bangla Sign Language (BdSL), particularly including both the letters and words, is limited. This paper assesses various models for...

The recognition of handwritten Bangla digits are important for automating Bengali language documents in applications like postal automation, banks cheques processing and so on. Due to the morphological variation between Bangla digits and Latin digits, the automatic recognition of Bangla digit system...

Sign language is an important form of communication for the hearing-impaired. Unfortunately, Bangla Sign Language (BdSL) has been largely neglected in the field of technology studies and applications. We presented a framework to incorporate deep learning into BdSL recognition to create an inclusive means...

Communication barriers between the Deaf and hard-of-hearing (DHH) community and the hearing population in Bangladesh persist due to the lack of automated Bengali Sign Language (BdSL) translation tools. This study proposes BdSL-Net, a real-time BdSL recognition framework based on computer vision and deep...

Finding the correct customer purchase prediction and crossapplicable segmentations will improve personalization and ultimately grow revenue within e-commerce. Current and traditional machine learning paradigms struggle to recognize the behavior of people as a non-linear and complex phenomenon that lacks...

The equitable distribution of government support including agricultural subsidies, disaster relief, educational scholarships, and elderly allowances remains a critical challenge in resource-constrained environments. This paper proposes a scalable, data-driven framework that integrates Machine Learning...

Specific determination of battery capacity is fundamental to enhance the safety, dependability and durability of electric vehicles (EVs). This paper presents a machine learning based data-driven approach to predicting the capacity, as determined by the State of Health (SOH) of lithium-ion batteries....

The role of resident halls on shaping personality traits among university students in Bangladesh has been investigated, focusing on sociability, emotional regulation, stress management and academic behavior. A dataset involving 226 students from several universities was analyzed based on different machine...

This study in Bangladesh is a more detailed analysis of how part time jobs affect Street CGPA of the students. The survey involved 611 Small-Island State respondents that included influences on employment type, % of average time per day spent and others differences. The research focuses on the nexus...

Maternal health complications are a major issue of concern in the world especially in developing countries. Early detection of the high-risk pregnancies is very important in minimizing the morbidity and mortality of the mother. The traditional clinical tests often rely on a few features and manual examination,...

Skin diseases are a major public health issue worldwide, requiring early diagnosis and intervention to prevent serious consequences, including melanoma-related deaths. However, the traditional dermatoscopic diagnosis of skin diseases is often time-consuming, subjective, and depends on the availability...

This research explored on how AI-based customer insights affect SI in Bangladeshi service organizations with the mediating role of KM. The study is based on the KBV and examines the effect of organizational capabilities, DAC, EC, ITI, and LS, on SI. Through purposive sampling, a quantitative method was...