ARAvijit Roy

Avijit Roy

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Avijit Roy has a total of 8 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Privacy-Preserving MIS Using Blockchain, Federated Learning and Zero-Trust Models

    by Avijit Roy, Emran Hossain, Hemayet Uddin Himel, Hasan Imam, Ruhul Amin Md Rashed, Harleen Kaur

  • Open Access

    Propasafe-Hybrid

    by Thomas Kimmeth, Avijit Roy, Vivek Sharma

    Propagandistic content increasingly circulates through online news and social media, where readers often encounter it with limited scrutiny, highlighting the need for reliable and fine‑grained detection. This paper introduces Propasafe‑Hybrid, a sentence‑level system that integrates a fine‑tuned transformer classifier with LLM‑based technique classification to identify, label, and explain specific propaganda strategies. The pipeline generates actionable outputs, including highlighted sentences, technique assignments, and concise rationales, so users can immediately understand why a sentence was flagged and how each label was determined. To control inference cost, Propasafe‑Hybrid employs a cost‑aware pre‑filtering stage that forwards only high‑likelihood sentences to LLMs, reducing token usage while preserving the underlying decision logic. Together, these design choices enhance the explainability, efficiency, and practical usability of sentence‑level propaganda detection in real‑world news environments.

  • Open Access

    Structural Silence: When AI Infrastructure Fails Speakers of Underrepresented Languages

    by Avijit Roy, Proma Roy

    Artificial intelligence tools for education and language support are increasingly framed as solutions to access gaps in under-resourced communities. Yet the infrastructure underlying these tools—training corpora, tokenization schemes, evaluation benchmarks, and deployment architectures—encodes a set of assumptions that systematically disadvantages speakers of underrepresented languages before a single model is trained. This paper examines those assumptions through the lens of Bengali, one of the world’s most widely spoken languages with roughly 285 million speakers (Ethnologue, 2025; International Communication and Leadership School, 2026), and the structural barriers that emerge when attempting to build AI-assisted educational tools for Bengali-speaking learners in low-connectivity environments. We identify four interlocking structural failures: a severe web presence gap (Bengali accounts for fewer than 0.5% of global web content despite representing nearly 4% of the global population) (Pimienta, 2024; W3Techs, 2026); a 67:1 training token deficit between English and Bengali in major multilingual corpora (Khan et al., 2024; Langlais et al., 2025); a tokenization penalty rooted in Bengali’s alphasyllabary script that compounds the data deficit by requiring higher token fertility rates (Shahriar and Barbosa, 2024); and a connectivity exclusion that renders cloud-dependent AI tools inaccessible to the rural populations most likely to benefit from them, where individual internet penetration stands at 36.5% compared to 71.4% in urban areas (Bangladesh Bureau of Statistics, 2025; The Daily Star, 2025). We argue that these failures are not isolated engineering problems awaiting technical fixes. They are the downstream consequences of longstanding resource allocation decisions, institutional priorities, and design defaults that did not center certain languages in mainstream AI development. Dataset scarcity is a structural barrier, not a researcher failure. Offline-first design is equity infrastructure, not a compromise. We close with specific directions for how the linguistics and AI communities might respond.