Research

Research publications, pre-publication papers, and working papers

Published

Martirosyan, V., Kamdar, R. (2026). "How Verification Mechanisms Alter Cultural Signals in Employer Reviews." Accepted to Americas Conference on Information Systems (AMCIS) 2026.

First author on a study analyzing how employer verification on Glassdoor affects the cultural signals employees communicate in their reviews. We used natural language processing to examine differences in review content between verified and unverified employers. Read on arXiv!
Read the 4 minute summary of the paper on gist.science here!

Prasad, K., Martirosyan, V., et al. (2025). "The Business Climate in Maryland and the Impact of Federal Policy Changes on Maryland Businesses: A Survey of Maryland Businesses." Research Report, Robert H. Smith School of Business, UMD.

Contributed to the article with data gathering, cleaning, preprocessing, and analysis. Article published.

Pre-Publication Papers

Colelough, B., Martirosyan, V., et al. (2026). "Reproducibility in Computer Science: An Empirical Multi-Year Replication Study of Neuro-Symbolic AI." Under review at IEEE ICSME 2026.

Co-first author, examining reproducibility in neuro-symbolic AI literature. We screened 5,497 papers from nine databases, narrowing the corpus to 1,365 candidates after deduplication and relevance filtering. I led quality control for the multi-stage reproduction protocol. arXiv pre-print coming soon.

Working Papers

Martirosyan, V., Jahani, E. (2026). "Misinformation Sharing Across Platforms and Its Determinants." In preparation.

First author, examining misinformation sharing behavior across social media platforms in India, Pakistan, and Nigeria. arXiv pre-print coming soon.

Martirosyan, V., Clark, J. "Machine Learning in Information Systems: A Benchmarking Framework." In preparation.

First author on a paper developing a comprehensive benchmarking framework for evaluating machine learning applications in information systems research. We provide standardized methodologies and metrics for comparing ML approaches across IS domains, as well as an AI-in-the-loop framework to batch-analyze papers. Manuscript in preparation.