iRepository (Institute of Business Administration)
Not a member yet
6878 research outputs found
Sort by
Optimistic oracle
No full textIshrat’s career is described in Sibtain Naqvi’s meticulously researched, well-structured biography Unravelling Gordian Knots: The Works and Worlds of Dr Ishrat Husain (2021). Dr Ishrat was perhaps too modest to write it himself. Naqvi is its surrogate parent. Ishrat’s family migrated from Agra in December 1947. His father established a legal practice in Hyderabad (Sindh), where Ishrat showed precocious promise by clearing his Matric exam at the age of 12. He graduated from Sindh University, then the University of British Columbia (Canada), after which he joined Pakistan’s Civil Service in 1964. That batch achieved a legendary reputation as an elite above an elite, occupying posts from the presidency of Pakistan downwards. In 1976, he obtained a US doctorate in economics and was recruited by the World Bank. It recognised his worth even when his government didn’t. He restructured chronic cases of national insolvency — Sierra Leone, Liberia, Ghana and finally Nigeria — and returned to the WB headquarters, a proven ‘wunderkind’. His talents were applied to other African countries and the emergent Central Asian Republics
Automated Product Filtering Using Large Language Models by Enhanced Similarity Detection in E-Commerce Catalog
No full textThis study aims to optimize the product listing process on e-commerce platforms by ensuring that all listings comply with social norms, government policies, and product listing guidelines. The project leverages automation to streamline the product listing procedure, reducing labor hours, minimizing the time required for new products to go live, and mitigating the risk of human error. The methodology involves utilizing Natural Language Processing (NLP) techniques to test product titles for prohibited keywords or non-dictionary terms that may be perceived as manipulations intended to bypass system checks. The process begins by uploading the product listing policy guidelines, followed by an evaluation to assess whether the product listings align with these policies. Any non-compliant products are then flagged. To evaluate product titles, the project employs cosine similarity with BERT embeddings to compare title words with a list of prohibited keywords, identifying high-risk titles. These titles are further analyzed and refined using Large Language Models (LLMs) such as GPT, which also generates corresponding descriptions. The revised titles are then tested against the listing policy to ensure compliance with the requirements for publication on the live site. Techniques such as regular expressions, cosine similarity, and fuzzy matching are employed to identify the highest-risk titles, effectively reducing operational costs in subsequent stages. Several pretrained text-to-text models, including OpenAI GPT- 4, GPT-4.0-min, Mistral, and Llama, are used for title correction and analysis. In the secondary phase, the system verifies product images to ensure they do not feature prohibited content. By generating a description of each image, the system compares it against the product listing policy to identify any prohibited items. This step helps detect instances where sellers may attempt to bypass quality control by submitting images that do not correspond with the product title. The system not only refines product titles and listings but also provides valuable insights into their compliance with listing policies. In the case of a policy violation, the system identifies the specific policy being violated and offers actionable feedback, ensuring that the listings are appropriate for publication on the live site
Telegram Agent for Cryptocurrency Analysis and Insights
No full textAs Large Language Models (LLMs) have grown in popularity and found applications across numerous domains, their potential for trading and financial analysis has also emerged. During the 2024 cryptocurrency bull run, AI agents gained significant traction, with tokens such as ai16z, AIXBT, and Virtuals Protocol experiencing substantial rises in value. These tokens provide access to agents that offer analysis and insights on other crypto tokens using blockchain data. In this project, I’ve developed an AI-driven agent suite on Telegram capable of enhancing cryptocurrency trading and analysis by leveraging advanced techniques: 1. Technical Analysis Agent: This agent performs real-time technical analysis of candle- stick charts for any cryptocurrency, across any timeframe. 2. Price Volume Tracker Agent: It evaluates a coin’s past performance by analyzing metrics such as trading volume, market capitalization, and price fluctuations. 3. Market Network Agent: Using a knowledge graph, this agent answers complex queries about the interrelationships between various cryptocurrencies, enabling traders to make informed decisions based on interconnected market dynamics. These agents aim to empower traders and analysts by offering deep insights, enabling faster and more accurate decision-making in the volatile cryptocurrency market. Keywords: Large Language Models, Retrieval Augmented Generation, Cryptocurrencies, Automated Trading, Technical Analysi
Urdu Sentence Boundary Detection with Statistical Learning Approaches
No full textSentence boundary detection (SBD) is a critical task in natural language processing (NLP), enabling accurate segmentation of text for downstream applications such as machine translation, summarization, and question answering. This project focuses on SBD for Urdu, a low-resource language with unique grammatical structures and a complex script. Our research focuses on transcribed Urdu text, which typically lacks punctuation and presents unique challenges in identifying sentence boundaries. This reflects real-world scenarios such as speech-to-text outputs and informal digital communication, where sentence segmentation must rely on linguistic features rather than punctuation cues. To tackle this challenge, this project explored multiple approaches. A rule-based method leveraging linguistic patterns and part-of speech (POS) tagging was used initially to understand the data and the nature of issues related to SBD. Statistical models, including Decision Trees, Random Forest, Logistic Regression, and XGBoost, utilized features such as XPOS and UPOS tags, among others. Additionally, deep learning models, including feedforward CNNs and LSTMs, were applied. Feature selection techniques were employed to optimize performance, and experiments were conducted to evaluate models on an imbalanced dataset with two target classes. The best model achieved an F-measure of 0.73 (73%) using XGBoost. Our findings reveal that the absence of punctuation poses significant challenges. Yet, meaningful improvements can be achieved through careful feature engineering and model selection. Statistical models demonstrated interpretability and efficiency. This research contributes to the growing body of work on low-resource languages and establishes a foundation for practical applications in Urdu NLP
Machine Learning Based Framework for Assessing Financial Institutions’ Liquidity
No full textThis project focuses on assessing financial and liquidity risk parameters in line with the Internal Liquidity Adequacy Assessment Process (ILAAP) and developing a machine learning-based framework to detect early signs of anomalies in liquidity risk assessment. The study analyzes financial data from five major banks in Pakistan over a ten-year period to identify key risk assessment cohorts, including Liquidity Coverage Ratio (LCR), Leverage Ratio (LR), Net Stable Funding Ratio (NSFR), Capital Adequacy Ratio (CAR), Return on Assets (ROA), and Advance to Core Deposit Funding. By applying machine learning models, the project aims to provide predictive insights into potential liquidity distress, enabling early intervention and improved risk management. The proposed framework automates the detection of warning signals in liquidity management, contributing to more robust financial risk governance and compliance with regulatory requirements. This project demonstrates how data-driven approaches can enhance liquidity risk management and strengthen the financial resilience of banking institution
Cricket and the City: A Spatial Exploration of the Factors Influencing Stadium Attendance in Karachi and Lahore
No full textThe purpose of this study was to identify the factors impact the cricket match attendance in the stadiums located in two of the largest urban centers in Pakistan, Karachi and Lahore. Despite cricket being a beloved sport nationwide, the turnover in the stadiums especially in the National Stadium, Karachi, has been particularly low and inconsistent over the course of matches. Accordingly, this research aims to understand if the limited attendance rate in the National Stadium is because of the lack of interest of the cricket fans in Karachi, as it is perceived on social media, or if it is the result of certain urban, infrastructural and social factors in the city by comparing it against the backdrop of Lahore’s Gadaffi Stadium. Accordingly, the research addresses the following primary question, What are the key factors affecting cricket match attendance at stadiums in Karachi and Lahore?” Hence, a mixed-method approach was employed for this research, which included quantitative data collected and spatial analysis using QGIS. The findings of the research indicate that contrary to popular belief, the low attendance in the National Stadium is not because of the disinterest in cricket. Instead, factors, such as traffic congestion, safety concerns, inefficient public transport, and lack of adequate stadium facilities dictate the decision of the residents to visit the stadium. Some of the experiences are also noted Lahore, however, owing to improved urban facilities, especially transport, Gadaffi Stadium is able to a higher attendance rate during cricket matches. The research calls for attention towards the amenities available in the stadiums and the urban facilities in each city that continue to impact the participation of residents in sports
IBA Karachi participates as Exclusive Partner at the 11th Deans and Directors Conference 2024
No full textDoes conventional, Islamic, and digital financial literacy augment financial well-being: the mediating role of financial behavior
No full textPurpose: This study aims to evaluate the mediating effect of financial behavior in the relationship between Conventional Financial Literacy, Islamic Financial Literacy, Digital Financial Literacy, and Financial Well-Being.
Methodology: A convenient sampling technique was used to collect data from 292 faculty members and Job holders to complete the study\u27s purpose.
Findings: In the study, the framework stresses the significance of financial behavior in mediating the link between Conventional Financial Literacy, Islamic Financial Literacy, Digital Financial Literacy, and Financial Well-Being. Additionally, it has been established that Conventional Financial Literacy, Islamic Financial Literacy, and Digital Financial Literacy, have a significant relationship with financial well-being and financial behavior.
Practical Implications: It is crucial to have a basic awareness of earnings, spending, and saving habits in order to contribute to a household\u27s financial well-being. Additionally, an encouraging economic strategy is required to oversee their financial well-being by offering employment, higher learning, instruction, pieces of training, and so forth. Offering fundamental financial instruction programs on money management, financial planning, financial assessment, risk-return diversification, and possibilities for investment should be part of efforts to promote conventional financial literacy, Islamic financial and digital financial literacy because they will help people develop the financial discipline that will help them achieve financial security.
Originality: To the best of our knowledge, this is the first study to consider Conventional Financial Literacy, Islamic Financial Literacy, Digital Financial Literacy, financial behavior, and Financial Well-Being merged in one model
Digitalization of Islamic Banking in Pakistan: An Exploratory Study of Change Management Perspective
No full textThe aim of this study is to explore challenges that are faced while implementing digitalization in an organization. The provision of banking or financial services using information technology is known as digital banking compared to online or mobile banking, it is a broader phrase. People resist to adopt new portals and system as these are difficult in the beginning due to new interface. People want to stick to the system that they are using since long as they have become habitual of It is seen that change is caused when organizations move for digitalization because their system adopt new IT structure that in itself bring quite new interfaces. This study is qualitative in nature since Interview were conducted from assistant vice presidents of digital banking group as the y were involved in the phase of transformation. Interviews were conducted up to the level of saturation. The collected data has been arranged into some usable information for analysis and thematically assessed when the necessary information was acquired from the in-depth interviews. In order to ensure that no important information provided by the respondents is missed, the interviews were also audio recorded with their consent. In order to establish a connection between the transcripts and produce comprehensive material from a synthesis of the interviews, they have been organized into categories according to the study questions in order to draw conclusions using thematic analysis