UIU Institutional Repository (United International University)
Not a member yet
3191 research outputs found
Sort by
A Hybrid Approach to Bangla Regional Text Classification Using BERT Ensemble and Region-Specific Lexical Oversampling
No full textpdfRegional text analysis reflects the lived realities of diverse communities by capturing
the linguistic richness and diversity present in various dialects. It bridges the gap between
everyday regional usage and standardized language forms, thereby enhancing the
inclusivity of language technologies. In this paper, we focus on five regional dialects in
Bangladesh, namely Chittagong, Sylhet, Noakhali, Barishal, and Rangpur, using a dataset
of 4,218 text samples. The dataset is validated by five regional experts and categorized into
three tiers based on an assigned agreement criterion. Tier 1 represents a strictly filtered,
high-confidence subset and is used primarily for evaluation. A set of region-specific special
words, which belong exclusively to their respective regions and are validated by domain
experts, is introduced. These words are used in a linguistically informed oversampling
technique to balance the dataset in both experiments. In the first experiment, we demonstrate
the effectiveness of the tiered dataset structure, where Tier 2 and Tier 3 (mediumand
low-confidence subsets) are used for training, and Tier 1 (high-quality subset) is used
for testing. In this setting, BanglaBERT achieves the best individual performance with
67.45% accuracy and a weighted F1-score of 67.62%. In the second experiment, we focus
exclusively on the Tier 1 dataset, applying a wide range of machine learning and deep
learning models to assess their effectiveness. The key contribution is a heterogeneous deep
ensemble technique that combines three BERT models, BanglaBERT, BUETBERT, and
DistilBERT, achieving an accuracy of 85.17% and a weighted F1-score of 84.84% on the
Tier 1 dataset.
iiiCSE UI
An Explainable Ensemble Convolutional Neural Network for Early Lung Cancer Prediction with Web Application
No full textCSE UIULung Cancer is one of the deadliest forms of cancer significantly contributing to the rising
mortality rates globally. The high mortality rate associated with lung cancer can largely be
attributed to its late detection, as symptoms often do not appear until the disease has reached
advanced stages. Early detection of anomalies in medical imaging, particularly at the initial stages,
is crucial for advancing both quantitative image analysis and patient care. In this context, our
research introduces a fully automated web application to predict lung cancer early on CT scan
images. This application leverages the power of a weighted average ensemble-based deep learning
framework that combines multiple neural network architectures to enhance the reliability of
automated classification. The work highlights the need for interpretability in clinical decision
support systems, going beyond classification. Our proposed approach is organized into three
independent phases. First, we performed image augmentation as part of the preprocessing stage
analyzing the IQ-OTH lung cancer dataset which implies that the model is trained on diverse and
enriched input data. The system incorporates ResNet50, VGG16, and a custom CNN, all of which
provide complementary feature-learning capabilities that improve overall predictive
dependability. The ensemble model clearly outperformed the individual networks, achieving an
accuracy of 92.28% on the IQ-OTH lung cancer dataset. The visual indications and regions that
most influence our model’s decisions are highlighted using Explainable Artificial Intelligence
techniques, particularly LIME and SHAP. By making the model’s inner workings more
understandable, these explanations hope to boost medical practitioners’ confidence.
Finally, the ensemble model was integrated into a user-friendly web application allowing users to
upload CT scan images, which are then analyzed to classify the images into normal, benign, or
malignant categories. Furthermore, our web application gives confidence levels for each prediction
increasing the credibility of its results. Our system provides medical practitioners with a smooth
and effective tool by integrating the precision of an ensemble model with the flexibility of a web
application.CSE UI
Recognition of Bangla and English Words in Bangla Texts Using a Modified BERT-base-NER Model
No full textCSE UIUA combination of Bangla and English words is commonly used, particularly on social media. This tendency greatly hampers the next generation’s ability to learn Bangla.
This study suggests an approach for identifying words in Bangla texts that are both
English and Bangla. This study also translates the identified English terms into standard Bangla words. The Transformer architecture, which uses an attention mechanism
to identify the connections between words and their contexts inside a text, is the foundation of bidirectional encoder representations from transformers (BERT). In this study, we
use the training input dataset to modify the BERT-base-NER model. For the name entity recognition (NER) task, the proposed BERT-base-NER model in this study achieves
state-of-the-art performance. For both the training and testing scenarios, we employ a
holdout cross-validation procedure. We used 80% of the entire data for training and 20%
for testing. We use the Google Translate API (application programming interface) to
translate the identified English words into standard Bangla words. In order to assess the
modified BERT-base-NER model, we applied the input dataset to the current machine
learning (ML) and deep learning (DL) techniques. Support vector machines (SVM) and
Naive Bayes (NB) are two components of the machine learning approach. Conversely, the
DL method uses bidirectional LSTM (BiLSTM), long short-term memory (LSTM), and
convolutional neural network (CNN). The improved BERT-base-NER model is highly accurate and efficient at identifying Bangla and English words, according to simulation data.
With an accuracy of 95%, the proposed BERT-base-NER model achieves the best result
among the current methods. For Bangla–English code-mixed text, this study presents a
reliable BERT-based word-level language identification system that successfully resolves
Banglish ambiguity and allows downstream Bangla language processing applications such
as standard Bangla conversion, machine translation, and information extraction.CSE UI
A Hybrid Approach to Bangla Regional Text Classification Using BERT Ensemble and Region-Specific Lexical Oversampling
No full textCSERegional text analysis reflects the lived realities of diverse communities by capturing
the linguistic richness and diversity present in various dialects. It bridges the gap between everyday regional usage and standardized language forms, thereby enhancing the
inclusivity of language technologies. In this paper, we focus on five regional dialects in
Bangladesh, namely Chittagong, Sylhet, Noakhali, Barishal, and Rangpur, using a dataset
of 4,218 text samples. The dataset is validated by five regional experts and categorized into
three tiers based on an assigned agreement criterion. Tier 1 represents a strictly filtered,
high-confidence subset and is used primarily for evaluation. A set of region-specific special
words, which belong exclusively to their respective regions and are validated by domain
experts, is introduced. These words are used in a linguistically informed oversampling
technique to balance the dataset in both experiments. In the first experiment, we demonstrate the effectiveness of the tiered dataset structure, where Tier 2 and Tier 3 (mediumand low-confidence subsets) are used for training, and Tier 1 (high-quality subset) is used
for testing. In this setting, BanglaBERT achieves the best individual performance with
67.45% accuracy and a weighted F1-score of 67.62%. In the second experiment, we focus exclusively on the Tier 1 dataset, applying a wide range of machine learning and deep
learning models to assess their effectiveness. The key contribution is a heterogeneous deep
ensemble technique that combines three BERT models, BanglaBERT, BUETBERT, and
DistilBERT, achieving an accuracy of 85.17% and a weighted F1-score of 84.84% on the
Tier 1 dataset.CSE UI
Impact of Management Information System in Recruiting in a Business Process Outsourcing (BPO) Organization - Quantanite Bangladesh
No full textDuring my last trimester at United International University for my "Bachelor of Business
Administration," I completed a four-month internship with Quantanite Bangladesh. Quantanite
Bangladesh is one of the country's leading BPO facilities. Business process outsourcing (BPO), on
the other hand, is the practice of a corporation assigning all of its internal tasks to another
organization. Business process outsourcing begins when an organization believes it can profit by
contracting with an outside firm to manage a business function, including supply chain
management, marketing, accounting, or management information systems.
In order to ensure client satisfaction, this study aims to investigate the quality of BPO services
and how Quantanite Bangladesh Limited manages and maintains them for its contractual
services. The report is structured as follows: Introduction, BPO industry overview, Company
Overview, MIS in HR and how it is helping to cut cost and make the process efficient. Lastly How
MIS process helps the IT department.
At Quantanite Bangladesh Limited, financial details were maintained through work experience.
Lastly, the information acquired from working with Quantanite Bangladesh Limited is used to
provide proposals for enhancing the field.
I became well acquainted with the company during my four-month internship and learned about
their role and importance. I learned a lot about various apps and received a lot of training relating
to my job. Working with these friendly coworkers and skilled lecturers also helped me get better
at Excel. I had the chance to participate greatly and carry out live job. As an intern, I can attest
that the entire journey was both alluring and challenging. I had the chance to test myself in every
manner during the course of these four months. I can thus say with confidence that one of the
best choices I've ever made was to collaborate with Quantanite Bangladesh
Organizational Activities of atB Jobs Bangladesh from Supply Chain Perspective
No full textThis report examines the operational activities of atB Jobs Bangladesh, a rapidly developing digital recruitment marketplace owned by atB Lab Ltd. It focuses on how the organization employs supply chain management (SCM) principles in its hiring process to improve efficiency, responsiveness, and service quality in Bangladesh's job market. The study explores atB Jobs ' operations through the lens of SCM, viewing recruiting as a service-based supply chain that includes data sourcing (job postings), processing (candidate matching), and distribution (shortlisting and employer handover). Job posting, screening, customer boarding, and cross-departmental collaboration are all key functions. The platform streamlines these procedures by utilizing AI, ATS, CRM tools, and resume databases. atB Jobs is mobile-first, user-friendly, and responsive to the needs of both employers
and job seekers. However, the report highlights issues such as AI matching limitations, infrastructure scalability, and the need for increased employer engagement. Recommendations include improving data quality, upgrading CRM systems, expanding
technological infrastructure, and boosting job seeker retention. From a supply chain standpoint, atB Jobs shows how SCM principles such as information flow, process integration, and capacity management may be applied to digital services. The internship demonstrated the platform's capacity to handle intangible resources, such as data and user engagement, with operational precision. In brief, atB Jobs successfully integrates recruitment and supply chain concepts. With a sustained focus on technology, stakeholder satisfaction, and process development, the platform is well-positioned for long-term success in Bangladesh's competitive job site business
Analysis of Consumers to Prevent Anti-Money Laundering & Combating Financial Terrorism: A Case Study of Dhaka Bank PLC
No full textThis report presents an in-depth analysis of the account opening process at Dhaka Bank PLC, focusing on compliance with Anti-Money Laundering (AML) and Combating the Financing of Terrorism (CFT) regulations. The account opening process is a critical function in banking, ensuring that new customers are onboarded securely while maintaining regulatory compliance.
The study highlights key components of the process, including customer verification, National ID (NID) validation, N-screening, sign scanning, customer transaction profiling, and forwarding accounts to the head office. These steps ensure that Dhaka Bank adheres to regulatory guidelines while minimizing risks associated with financial crimes.
In Dhaka Bank PLC, compliance with Bangladesh Bank regulations and global AML/CFT policies is an essential aspect of the account opening process. Strict due diligence measures are followed to verify customer authenticity, prevent fraudulent activities, and maintain transparency in financial transactions. This includes thorough risk assessments and the integration of technological advancements to streamline verification and documentation procedures.
In conclusion, the account opening process at Dhaka Bank PLC is structured and well-regulated, ensuring efficiency and security while maintaining compliance with AML and CFT guidelines. The study identifies areas for improvement, such as enhancing automation and digital verification tools, to further optimize the process. By maintaining strict regulatory standards and leveraging technology, Dhaka Bank PLC continues to uphold the integrity of its banking operations and contribute to a more secure financial system
Customer Centricity in the Marketing Communication Industry: A Case of PASSWORD COMMUNICATION
No full textThis internship report provides an in-depth analysis of the operations, challenges, and opportunities within Password Communication, a full-service marketing agency in Bangladesh, where I serve as the Founder & CEO. The agency specializes in ATL and BTL marketing, digital marketing, web and mobile application development, SMS and email marketing, TVC/OVC production, and logistics support.
The report begins with an overview of the marketing industry in Bangladesh, highlighting its rapid evolution driven by digital transformation, shifting consumer behavior, and increasing competition. It also explores external factors such as political, legal, and technological influences and provides a detailed assessment of industry rivalry.
The second part focuses on the organizational structure and business model of Password Communication. It includes a review of the agency's mission, service portfolio, customer base, and operational facilities. A SWOT analysis is conducted to identify strengths, weaknesses, opportunities, and threats, followed by strategic recommendations to address these elements effectively.
The report also emphasizes the agency’s strong commitment to customer centricity, demonstrated through tailored service delivery, transparent communication, and continuous feedback integration. Key internal and external factors influencing this customer-first culture are explored, including leadership, technology, client diversity, and market competition.
Lastly, the report presents insights from the internship experience itself, reflecting on practical learning, organizational dynamics, and areas for future development. It concludes with a set of recommendations aimed at reinforcing Password Communication’s position in the dynamic marketing landscape of Bangladesh
Internship Report on General Banking Operations of City Bank PLC & Customer Perception of Citytouch App
No full textThis report will aim to provide a detailed overview of the general banking operation and digital transformation of City Bank. This paper starts with an overview of the bank’s background and the variety of products and services it offers. Then I did a short industry analysis. This report will give you an idea about the branch operations activities. A major focus of the report is its digital products and finding out the current state of “Citytouch”, the biggest digital banking platform of the bank. A survey was conducted with 64 users to gather their experiences with the app. The result shows why it is the core digital banking platform of City Bank. This report will also provide you with information about other digital products of the bank. Additionally, this report will also share my personal internship experiences. Lastly, based on research and user feedback, the report concludes with specific recommendations aimed at improving service quality and keeping customer satisfaction high
An Email Marketing Plan for an Online Startup Company
No full textThis report, “A Project Report on an Email Marketing Strategy for an Online Startup Company,” explores the increasing significance of email marketing for startups functioning in the current competitive digital environment. It emphasizes how email marketing acts as a strong and affordable method to enhance consumer interaction, create leads, and cultivate brand loyalty.
Key topics examined strategic campaign components—such as subject lines, content layout, and calls to action—along with the effects of personalization, segmentation, and automation platforms like Mailchimp, HubSpot, and Klaviyo. The report additionally assesses performance indicators like open rates, click-through rates, and conversions to measure total effectiveness.
A case study on Yeppoonie Smoothie demonstrates practical application, highlighting the use of Klaviyo for automated flows, strategies for list building, A/B testing, and branding design. The report highlights email marketing as a significant factor in Yeppoonie’s customer engagement and sales increase, while also recognizing typical challenges like restricted startup funding, data privacy laws, and competition in inboxes.
The report wraps up with practical suggestions, such as diversifying suppliers, scalable investments, localized content, loyalty programs, and boosting user-generated content—highlighting the enduring importance of a data-driven, ethically responsible email marketing approach for startup growth and sustainability