About

Irena Vodenska is a Professor of Finance at Boston University’s Metropolitan College, where she also serves as the Director of Finance Programs and Chair of Administrative Sciences. Her research focuses on network theory and complexity science in macroeconomics, conducting both theoretical and applied interdisciplinary research using quantitative approaches. Her work models interdependences of financial networks, banking system dynamics, and global financial crises, with particular attention to early warning indicators and systemic risk propagation across interconnected financial and economic networks. She studies the effects of news announcements on financial markets, corporations, and global economic systems, utilizing neural networks and deep learning methodologies for natural language processing to analyze factors influencing corporate performance and economic trends.

Research topics

• Computer Science
• Artificial Intelligence
• Economics
• Finance
• Business
• Computer Security
• Financial system
• Natural Language Processing
• Political Science
• Actuarial science
• Machine Learning
• Natural resource economics
• Distributed computing
• Monetary economics
• Macroeconomics
• Engineering
• Risk analysis (engineering)
• International economics
• Reliability engineering
• Microeconomics

Selected publications

• When AI Fails, What Works? A Data-Driven Taxonomy of Real-World AI Risk Mitigation Strategies

  ArXiv.org · 2026-03-04

  articleOpen access

  Large language models (LLMs) are increasingly embedded in high-stakes workflows, where failures propagate beyond isolated model errors into systemic breakdowns that can lead to legal exposure, reputational damage, and material financial losses. Building on this shift from model-centric risks to end-to-end system vulnerabilities, we analyze real-world AI incident reporting and mitigation actions to derive an empirically grounded taxonomy that links failure dynamics to actionable interventions. Using a unified corpus of 9,705 media-reported AI incident articles, we extract explicit mitigation actions from 6,893 texts via structured prompting and then systematically classify responses to extend MIT's AI Risk Mitigation Taxonomy. Our taxonomy introduces four new mitigation categories, including 1) Corrective and Restrictive Actions, 2) Legal/Regulatory and Enforcement Actions, 3) Financial, Economic, and Market Controls, and 4) Avoidance and Denial, capturing response patterns that are becoming increasingly prevalent as AI deployment and regulation evolve. Quantitatively, we label the mitigation dataset with 32 distinct labels, producing 23,994 label assignments; 9,629 of these reflect previously unseen mitigation patterns, yielding a 67% increase of the original subcategory coverage and substantially enhancing the taxonomy's applicability to emerging systemic failure modes. By structuring incident responses, the paper strengthens "diagnosis-to-prescription" guidance and advances continuous, taxonomy-aligned post-deployment monitoring to prevent cascading incidents and downstream impact.

  PublisherOA PDF

• When AI Fails, What Works? A Data-Driven Taxonomy of Real-World AI Risk Mitigation Strategies

  Open MIND · 2026-03-04

  preprint

  Large language models (LLMs) are increasingly embedded in high-stakes workflows, where failures propagate beyond isolated model errors into systemic breakdowns that can lead to legal exposure, reputational damage, and material financial losses. Building on this shift from model-centric risks to end-to-end system vulnerabilities, we analyze real-world AI incident reporting and mitigation actions to derive an empirically grounded taxonomy that links failure dynamics to actionable interventions. Using a unified corpus of 9,705 media-reported AI incident articles, we extract explicit mitigation actions from 6,893 texts via structured prompting and then systematically classify responses to extend MIT's AI Risk Mitigation Taxonomy. Our taxonomy introduces four new mitigation categories, including 1) Corrective and Restrictive Actions, 2) Legal/Regulatory and Enforcement Actions, 3) Financial, Economic, and Market Controls, and 4) Avoidance and Denial, capturing response patterns that are becoming increasingly prevalent as AI deployment and regulation evolve. Quantitatively, we label the mitigation dataset with 32 distinct labels, producing 23,994 label assignments; 9,629 of these reflect previously unseen mitigation patterns, yielding a 67% increase of the original subcategory coverage and substantially enhancing the taxonomy's applicability to emerging systemic failure modes. By structuring incident responses, the paper strengthens "diagnosis-to-prescription" guidance and advances continuous, taxonomy-aligned post-deployment monitoring to prevent cascading incidents and downstream impact.

  DOI

• Complex Hilbert Principal Component Analysis Approach To Understand Economic Policy Uncertainty and Geopolitical Risk

  SSRN Electronic Journal · 2025-01-01

  preprintOpen access1st authorCorresponding
  PublisherDOI

• Navigating Global ESG Investment Regulations Using AI

  WORLD SCIENTIFIC eBooks · 2024-08-20

  book-chapter1st authorCorresponding

  Global initiatives such as the Principles for Responsible Investment (PRI) or the UN Sustainable Development Goals (SDGs) are well poised to catalyze the transition toward climate justice, food security, and agricultural resilience. These initiatives motivate countries to develop frameworks for sustainability reporting to facilitate the transition economy. Globally, policymakers and regulators have proposed regulations to increase the transparency and uniformity of reporting surrounding climate change and corporate social responsibility. We investigate over 100 global documents related to regulatory developments in sustainable taxonomies, climate disclosures, and Environmental, Social, and Governance (ESG) fund requirements reported in the 2023 Sustainable Fitch Tracker of ESG Regulations and Reporting Standards. We propose a model utilizing the power of Artificial Intelligence and Large Language Models to analyze global regulatory documents to capture sustainability-related risks and opportunities defined by the Sustainable Accounting Standard Board (SASB). We compare the performance of keyword-based and ChatGPT-based models and find that the ChatGPT model successfully detects a greater number of global regulatory documents containing the SASB topics. Our results show that the European Union is the leader in having the largest amount of effective and mandatory regulatory coverage of SASB categories, followed by Nigeria and Saudi Arabia. The United States is the leader in effective and non-mandatory regulatory coverage, followed by Malaysia and Mexico.

  PublisherDOI

• Understanding Worldwide Natural Gas Trade Flow for 2017 to 2022: A Network-Based Approach

  Communications in computer and information science · 2024-01-01 · 2 citations

  book-chapter
  PublisherDOI

• Comparing the performance of ChatGPT and state-of-the-art climate NLP models on climate-related text classification tasks

  E3S Web of Conferences · 2023-01-01 · 8 citations

  articleOpen accessSenior author

  Recently, there has been a surge in general-purpose language models, with ChatGPT being the most advanced model to date. These models are primarily used for generating text in response to user prompts on various topics. It needs to be validated how accurate and relevant the generated text from ChatGPT is on the specific topics, as it is designed for general conversation and not for context-specific purposes. This study explores how ChatGPT, as a general-purpose model, performs in the context of a real-world challenge such as climate change compared to ClimateBert, a state-of-the-art language model specifically trained on climate-related data from various sources, including texts, news, and papers. ClimateBert is fine-tuned on five different NLP classification tasks, making it a valuable benchmark for comparison with the ChatGPT on various NLP tasks. The main results show that for climate-specific NLP tasks, ClimateBert outperforms ChatGPT.

  PublisherOA PDFDOI

• Interdependencies Between Cryptocurrency Markets, Precious Metals and Energy Resources

  Lecture notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering · 2023-01-01 · 1 citations

  book-chapter
  PublisherDOI

• Post-COVID-19 depression prediction using Twitter data

  OpenBU (Boston University) · 2023-01-01

  other

  Accepted manuscript

  Publisher

• Using ML and Explainable AI to understand the interdependency networks between classical economic indicators and crypto-markets

  Physica A Statistical Mechanics and its Applications · 2023-05-19 · 7 citations

  article
  PublisherDOI

• A New Look at Calendar Anomalies: Multifractality and Day-of-the-Week Effect

  Entropy · 2022-04-17 · 8 citations

  articleOpen accessCorresponding

  Stock markets can become inefficient due to calendar anomalies known as the day-of-the-week effect. Calendar anomalies are well known in the financial literature, but the phenomena remain to be explored in econophysics. This paper uses multifractal analysis to evaluate if the temporal dynamics of market returns also exhibit calendar anomalies such as day-of-the-week effects. We apply multifractal detrended fluctuation analysis (MF-DFA) to the daily returns of market indices worldwide for each day of the week. Our results indicate that distinct multifractal properties characterize individual days of the week. Monday returns tend to exhibit more persistent behavior and richer multifractal structures than other day-resolved returns. Shuffling the series reveals that multifractality arises from a broad probability density function and long-term correlations. The time-dependent multifractal analysis shows that the Monday returns' multifractal spectra are much wider than those of other days. This behavior is especially persistent during financial crises. The presence of day-of-the-week effects in multifractal dynamics of market returns motivates further research on calendar anomalies for distinct market regimes.

  PublisherOA PDFDOI

Recent grants

• EAGER: Modeling systemic risk: Finding precursors of emerging financial crises

  NSF · $58k · 2014–2015

Frequent coauthors

• Lou Chitkushev

  Boston University

  30 shared

• Dimitar Trajanov

  Boston University

  28 shared

• H. Eugene Stanley

  Boston University

  25 shared

• Shlomo Havlin

  Bar-Ilan University

  15 shared

• Alexander P. Becker

  Boston University

  14 shared

• Hideaki Aoyama

  RIKEN

  13 shared

• Sergey V. Buldyrev

  Yeshiva University

  13 shared

• Shlomo Havlin

  Bar-Ilan University

  12 shared

Education

• PhD

  Boston University

Awards & honors

• Interdisciplinary research grants awarded by the European Co…
• Research grants awarded by the Network Science Division of t…
• Research grants awarded by the National Science Foundation (…