An Empirical Study of Gendered Stereotypes in Emotional Attributes for Bangla in Multilingual Large Language Models

Jayanta Sadhu (BUET), Maneesha Rani Saha, Dr. Rifat Shahriyar (Professor, BUET)

Abstract: The influence of Large Language Models (LLMs) is rapidly growing, automating more jobs over time. Assessing the fairness of LLMs is crucial due to their expanding impact. Studies reveal the reflection of societal norms and biases in LLMs, which creates a risk of propagating societal stereotypes in downstream tasks. Many studies on bias in LLMs focus on gender bias in various NLP applications. However, there's a gap in research on bias in emotional attributes, despite the close societal link between emotion and gender. This gap is even larger for low-resource languages like Bangla. Historically, women are associated with emotions like empathy, fear, and guilt, while men are linked to anger, bravado, and authority. This pattern reflects societal norms in Bangla-speaking regions. We offer the first thorough investigation of gendered emotion attribution in Bangla for both closed and open source LLMs in this work. Our aim is to elucidate the intricate societal relationship between gender and emotion specifically within the context of Bangla. We have been successful in showing the existence of gender bias in the context of emotions in Bangla through analytical methods and also show how emotion attribution changes on the basis of gendered role selection in LLMs. All of our resources including code and data are made publicly available to support future research on Bangla NLP.

LLM Emotion Attributes Bangla Gender Bias

  

Social Bias in Large Language Models For Bangla: An Empirical Study on Gender and Religious Bias

Jayanta Sadhu (BUET), Maneesha Rani Saha, Dr. Rifat Shahriyar (Professor, BUET)

Abstract: The rapid growth of Large Language Models (LLMs) has put forward the study of biases as a crucial field. It is important to assess the influence of different types of biases embedded in LLMs to ensure fair use in sensitive fields. Although there have been extensive works on bias assessment in English, such efforts are rare and scarce for a major language like Bangla. In this work, we examine two types of social biases in LLM generated outputs for Bangla language. Our main contributions in this work are: (1) bias studies on two different social biases for Bangla (2) a curated dataset for bias measurement benchmarking (3) two different probing techniques for bias detection in the context of Bangla. This is the first work of such kind involving bias assessment of LLMs for Bangla to the best of our knowledge. All our code and resources are publicly available for the progress of bias related research in Bangla NLP.

LLM Bangla Gender Bias Bangla Religious Bias Ethics Fairness

  

Dependency, Deadline and Priority Aware Multi-Queue Dynamic Task Scheduling Using Heterogeneous Resources in Fog Environment

Muhammad Ehsanul Kader (BUET)*, Maneesha Rani Saha*, Dr. Rezwana Reaz (Asistant Professor, BUET) (* equal contribution)

Undergraduate Thesis

Abstract: Internet of Things (IoT) has emerged as a transformative paradigm and revolutionized our way of engaging with technology in everyday life. The limited computing resources and storage capacity of IoT devices have necessitated the integration of heterogeneous fog resources in IoT task execution. Moreover, some IoT applications such as healthcare services are highly delay-sensitive; hence need immediate processing and response. On the other hand, the best utilization of heterogeneous storage and processing capacity of fog resources plays a pivotal role in improving the response time of IoT applications. Thus, an adaptive task scheduling algorithm for executing IoT tasks in a fog environment is required which will consider the heterogeneity of available resources and various response time requirements of different IoT applications. In this paper, we propose a non-preemptive batch task scheduling algorithm with dynamic resource allocation. The proposed method addresses the challenges of scheduling tasks with varying priority levels, deadline constraints, and dependencies utilizing the concepts of Directed Acyclic Graphs (DAGs), priority queues, and dynamic queue switching. We introduce a novel numerical scoring mechanism that prioritizes the execution of tasks with higher urgency and critical dependencies. The major goals of the proposed technique are to reduce the overall average response time, makespan, and average deadline violation time; and increase the throughput and task completion rate. We have implemented our algorithm in a Java based proprietary simulator. Our simulation results demonstrate that the proposed algorithm significantly improves the average response time, makespan, throughput, task completion rate and deadline violation time compared to the existing priority based batch scheduling algorithms. Precisely, our algorithm reduces average response time by at least 50.2% and makespan by at least 8.18%, and increases throughput by at least 13% and task completion rate by at least 4.3% compared to Shortest Job First (SJF) and Multi Level Feedback Queue (MLFQ).

IoT Task Scheduling Directed Acyclic Graphs Multi Level Priority Queue Fog Computing Heterogeneous Systems Dynamic Resource Allocation