In this work we are in pursuit of exploring the potential of existing LLMs to guide non-native english speaking children to improve their proficiency in english writing. I am developing a mobile app to be used as the ChatGPT supported EFL learning platform.

Field: Human-Computer Interaction and Cybersecurity

• SupervisorDr Sameer Patil, Associate Professor, KSoC and Faculty Co‑Director, DATASET Initiative at the University of Utah

In this work we are in pursuit of exploring the potential of existing LLMs to guide non-native english speaking children to improve their proficiency in english writing. I am developing a mobile app to be used as the ChatGPT supported EFL learning platform.

Field: Human-Computer Interaction

• SupervisorDr. A. B. M. Alim Al Islam, Professor, CSE, BUET
• IdeatorMahjabin Nahar, Graduate Teaching Assistant at Penn State. Penn State College of Information Sciences and Technology
• Collaborative Work With Penn State University Under the Supervision Of Dongwon Lee, Ph.D. Penn State University, USA
• Dev Tools and Frameworks React-native, Django, Firebase, Twillo, ChatGPT API

Short videos or reels, a new form of entertainment, have become quite popular among the users of Social Networking Sites and have considerably boosted the user engagement. However, some recent studies have shown that frequent context switching caused by short videos can have detrimental effects on human cognition. In this work, we have explored the impact of short videos on human cognition and also proposed a novel approach for content recommendation in order to reduce their detrimental effects. This work is currently under review in a renowned conference.

Field: Human-Computer Interaction, Recommendation Systems

• SupervisorDr. Sadia Sharmin, Associate Professor, CSE, BUET

In this work, we have investigated the awareness and security concerns of recommendation system users. We are interviewing participants and analyzing the data using thematic analysis.

Field: Human-Computer Interaction, Recommendation Systems

• SupervisorDr. Sadia Sharmin, Associate Professor, CSE, BUET
• Tools for Quantitative Analysis SmartPLS

In this work, we are investigating the privacy perspectives of shared mobile phones within households. I have interviewed 13 participants. Currently we are in the process of more data collection and analysis of the data using thematic analysis.

Field: Human-Computer Interaction

• SupervisorDr. A. B. M. Alim Al Islam, Professor, CSE, BUET
• IdeatorRudaiba Adnin, Ph.D. Student at Northeastern University

Undergraduate Thesis
In this work, I have uncovered the underlying theory of the improved attack performance and undetectability of a knowledge graph based state of the art shilling attack model. I have implemented 3 different types of recommendation systems, 6 types of Shilling Attacks (including 1 obfuscation) and 3 benchmark shilling attack detection methods and evaluated their performance on the renowned MovieLens and YahooMovies datasets.

Field: Recommendation System, Machine Learning, Security

• SupervisorDr. Mahmuda Naznin, Professor, CSE, BUET
• Dev Tools and Frameworks Python, NumPy, Pandas
• Thesis Defense Slide