Research Reports on Computer Science

Exploring Sentiment Analysis: A Study on Rheumatoid Arthritis and Lupus in Healthcare

Reddy Sowmya Vangumalla — 2024-08-13

Patients with autoimmune disorders such as lupus and rheumatoid arthritis (RA) have significant life-changing effects on both their physical and mental health. Using patient testimonies collected from social media and internet forums, this paper does a thorough investigation. Using natural language processing methods, we analyze textual data to reveal patients' common attitudes, feelings, worries, and coping mechanisms. Our goal is to give a comprehensive understanding of the emotional aspects of having an autoimmune disease, which will help researchers, support groups, and medical professionals better meet the psychosocial requirements of their patients. We also examine scholarly works published between 2019 and 2024, which deepens our comprehension of the affective dimensions of these situations. Through close examination of text data, we are able to identify common attitudes, feelings, worries, and coping mechanisms among patients. Our research aims to provide useful information to researchers, healthcare providers, and support groups to improve the way psychological requirements in autoimmune disorders are managed. Finally, the challenges of sentiment analysis are examined in order to define future directions.

A Comparative Analysis of Model Agnostic Techniques for Explainable Artificial Intelligence

Yifei Wang — 2024-08-07

Explainable Artificial Intelligence (XAI) has become essential as AI systems increasingly influence critical domains, demanding transparency for trust and validation. This paper presents a comparative analysis of prominent model agnostic techniques designed to provide interpretability irrespective of the underlying model architecture. We explore Local Interpretable Model-agnostic Explanations (LIME), SHapley Additive exPlanations (SHAP), Partial Dependence Plots (PDP), Individual Conditional Expectation (ICE) plots, and Anchors. Our analysis focuses on several criteria including interpretative clarity, computational efficiency, scalability, and user-friendliness. Results indicate significant differences in the applicability of each technique depending on the complexity and type of data, highlighting SHAP and LIME for their robustness and detailed output, whereas PDP and ICE are noted for their simplicity in usage and interpretation. The study emphasizes the importance of context in choosing appropriate XAI techniques and suggests directions for future research to enhance the efficacy of model agnostic approaches in explainability. This work contributes to a deeper understanding of how different XAI techniques can be effectively deployed in practice, guiding developers and researchers in making informed decisions about implementing AI transparency.

A Novel Optimization Algorithm for Otsu's Entropy-Based Multi-Level Thresholding for Image Segmentation

Karri Chiranjeevi — 2024-08-05

In applications involving image processing, segmentation is an essential stage. This procedure divides the image's pixels into various classes, enabling the examination of the scene's objects. Finding the ideal collection of thresholds to correctly segment each image is a challenge that multilevel thresholding solves with ease. The optimal thresholds can be found using methods like Otsu's between-class variance or kapur's entropy, but they are computationally costly when there are more than two thresholds. This study presents a novel meta-heuristic algorithm, Election-Based Optimization Algorithm (EBOA) to discover the optimal threshold configuration with Otsu as the objective function, to solve this kind of problem. The obtained results proved better in WPSNR, PSNR, SSIM, FSIM and misclassification error and segmented image quality when compared with existing algorithms.