Projects | Ravinithesh

US 2020 elections through the eyes of TikTok

This study tried to comprehend the political communication about the US presidential election on TikTok. A set of videos that contained the hashtags from a predetermined set were collected and analyzed. The results suggest that the communication is interactive, and the young people are talking about issues rather than parties or personalities. However, they season their message with fun and trendy challenges.

Hollywood and Pentagon Relationship

This project investigated the relationship between the Department of Defense of the United States of America and Hollywood. The results show that Hollywood receives the support of costly and military-related paraphernalia from the top agencies of the DoD and in return sheds good light on DoD in the movies. Simultaneously, DoD cast off the movies that narrated uncomfortable stories about the United States.

Structural Similarities And Differences In Chinese Folk Songs

This study explored the structural differences in Chinese Folk Songs and compared their similarity across various geographical regions. The analysis of scale degrees distributions and intervals reveal deep cross-regional similarities. However, the pitch transition asymmetries with respect to the metrical weights show that some regions clearly distinguish from the others. Overall, the findings highlight the directedness of music across all the examined regions.

Face of the War

This project presents a web-based interactive visualization of photographs of soldiers who took part in the Great War. While presenting the portraits of soldiers along with their records, the visualization manifests the question, who is the real face of the war?

Terzani Online Museum

Through this project, we try to bring the photographs taken during Mr. Terzani's trip to South Asian countries to an online platform and make a large collection of photos navigable and searchable.

Adaptive Learning in Extreme Learning Machines

Developed formulations and software for adaptive learning (using Actor Critic) in Extreme Learning Machines (ELM) and it performed 2X faster and at comparable level with a similar system developed used Artificial Neural Networks on a hard industrial problem.