Sriram Karthik

Sriram Karthik


Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
14 October 2025 in Preprints.org
Verified
Preprint
Open access
PDF file
GSGrigory Statnikov
Grigory Statnikov
SKSriram Karthik
Sriram Karthik
YBYoussef Boumhaout
Youssef Boumhaout
AEAbdelrahman Elfikky
Abdelrahman Elfikky
This survey paper presents a comprehensive review of convolution operations in Convolutional Neural Networks (CNNs). We trace the evolution of convolutions from their mathematical foundations to their implementation in modern deep learning architectures. We examine various types of convolutions, their properties, and their applications across different domains, including image processing, audio analysis, natural language processing, and scientific applications. Additionally, we discuss the biological inspiration behind CNNs, key architectural innovations, and emerging trends in the field. This paper serves as a resource for researchers and practitioners interested in understanding the fundamental building blocks of CNNs and their continuing evolution.

17 October 2025 in Preprints.org
Verified
Preprint
Open access
PDF file
SKSriram Karthik
Sriram Karthik
SKSkandhan Karthik
Skandhan Karthik
In this study, we analyze the sentiment and thematic dynamics of Twitter and Reddit—two distinct social media ecosystems—from a natural language processing perspective. We compiled and examined a dataset of 300 posts from each platform, all centered on AI responsibility and ethics. Sentiment Patterns were first assessed using the VADER (Valence Aware Dictionary and Sentiment Reasoner) tool, a lexicon, and a rule- Based sentiment analyzer optimized for social media. To com- To complement this, we employed BERTopic, a machine learning–based topic modeling framework that leverages trans- former embed- dings and HDBSCAN clustering to detect and visualize recurring themes in discourse. Cosine similarity on topic centroids was used to map semantic relationships, while Euclidean distance was applied during clustering to ensure robust separation of themes. Our results show a striking difference: Twitter exhibits greater emotional variation and polarization, with sentiment scores frequently reaching both positive and negative extremes, whereas Reddit maintains a more consistent, balanced tone centered near neutrality. This difference emerged consistently across multiple visualization methods, including sentiment heatmaps, box plots, violin plots, radar charts, and topic similarity maps. By combining sentiment scoring with topic mapping, our work highlights how both emotional tone and the thematic framing of AI ethics unfold differently across online spaces, offering insights for public opinion measurement, platform governance, and algorithmic responsibility. The insights we gainged from this comparitive analysis had showed us how public opinion’s were being measured, platform governance, and algorithimc responsiblity in the context of social media discourse.