About

Liye Ma is a Professor of Marketing at the Robert H. Smith School of Business at the University of Maryland. He is an expert on AI/Machine Learning and Digital Marketing. His research focuses on the dynamic interactions of consumers and firms on Internet, social media, and mobile platforms. He develops statistical, econometric, and machine learning methods to analyze the drivers of consumer actions in the digital economy, and uses the findings to help companies develop digital marketing strategies and optimize marketing decisions. Dr. Ma has published in leading academic journals and serves on editorial review boards. He has received research grants from notable institutions and has been recognized as a finalist for prestigious awards. At the Smith School, he teaches courses related to Big Data, AI, Data Science, and Doctoral Seminars.

Research topics

• Computer Science
• Artificial Intelligence
• Marketing
• Information Retrieval
• Business
• Data Mining
• Machine Learning
• Reliability engineering
• Knowledge management
• Advertising
• Engineering
• Mathematics
• Mathematical optimization
• Data science

Selected publications

• NQO1 phase condensation promotes stress granule assembly to facilitate pancreatic carcinogenesis

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-09-29

  preprintOpen access

  Abstract G3BP1 promotes pancreatic ductal adenocarcinoma (PDAC) tumorigenesis driven by oncogenic KRAS mutations through liquid-liquid phase separation (LLPS)-mediated assembly of stress granules (SGs). However, the regulatory mechanisms remain elusive. We identify the antioxidant enzyme NAD(P)H quinone dehydrogenase 1 (NQO1) as a novel SG regulator that enhances SG assembly in pancreatic cancer cells independently of reactive oxygen species (ROS). Mechanistic studies reveal that NQO1 does not regulate the expression of G3BP1 or other SG-associated genes. Instead, NQO1 undergoes LLPS dependent on its RNA-binding K homology (KH) like domain. Further analysis demonstrates that residues 121–131 of NQO1 directly interact with G3BP1’s RNA-binding domain (RBD), enhancing the multivalency of G3BP1 complexes to potentiate LLPS-driven SG assembly. This interaction accelerates cell proliferation, KRAS mutation-induced acinar-to-ductal metaplasia (ADM), and pancreatic carcinogenesis. Notably, the interaction between NQO1 residues 121–131 and G3BP1’s RBD is essential for NQO1 phase condensation under stress condition. Integrative analysis of human PDAC transcriptomic datasets reveals a weak association between NQO1 and G3BP1 levels. Importantly, both NQO1 and G3BP1 formed biological condensates and co-localized in the lesions of human PDAC tissue sections. Our study uncovers a novel KRAS mutation-driven mechanism of pancreatic carcinogenesis through the lens of phase separation, transcending conventional gene expression regulation and offering new insights into non-canonical KRAS-NQO1-G3BP1-SG regulatory networks in PDAC initiation.

  PublisherOA PDFDOI

• Program-Ad Congruency in TV Advertising: A Multi-Modal Analysis of Thematic and Emotional Congruency

  SSRN Electronic Journal · 2025-01-01

  preprintOpen access
  PublisherDOI

• Picture It or Type It: Comparing Visual and Textual Search in E-Commerce

  SSRN Electronic Journal · 2025-01-01

  preprintOpen accessSenior author
  PublisherDOI

• The value of geolocation information across the promotion funnel

  Marketing Letters · 2025-11-25

  articleSenior author
  PublisherDOI

• Predicting Purchase Intent: Deciphering Customer Interactions with AI Assistants

  SSRN Electronic Journal · 2024-01-01 · 2 citations

  articleOpen access
  PublisherDOI

• Application of Global Importance Measures in RISMC for Dynamic Risk Assessment

  2024-08-04

  article1st authorCorresponding

  Abstract Risk-Informed Safety Margin Characterization (RISMC) integrates traditional deterministic safety analysis methods with probabilistic safety analysis approaches by employing stochastic sampling of uncertain parameters and dynamic branching simulations of nuclear power plants. It enables time-dependent, uncertainty-aware, and multi-level risk assessments. In the process of RISMC analysis, quantifying the impacts of uncertainty parameters, response events, and system components on nuclear power plant risk can support the implementation of risk analysis results in nuclear power plants. However, traditional importance measurement methods cannot directly apply to RISMC. This study adopts the global importance measurement method to measure the time-dependent importance. The selected global importance measurement method considers input parameter uncertainties and uncertainties from interactions between input parameters. It is demonstrated on Station Blackout (SBO) accidents modeled with dynamic event trees and compared with traditional importance metrics used in probabilistic risk assessment. The risk importance of systems and physical design parameters in SBO accidents is ranked. Results indicate timely AC power recovery significantly reduces risk, while the importance of TAFW failure diminishes over time. Future research is needed to investigate the discrepancies between the global importance measurement method and traditional methods.

  PublisherDOI

• How Does Competition Affect Exploration vs. Exploitation? A Tale of Two Recommendation Algorithms

  Management Science · 2023-03-30 · 27 citations

  article

  Through repeated interactions, firms today refine their understanding of individual users’ preferences adaptively for personalization. In this paper, we use a continuous-time bandit model to analyze firms that recommend content to multihoming consumers, a representative setting for strategic learning of consumer preferences to maximize lifetime value. In both monopoly and duopoly settings, we compare a forward-looking recommendation algorithm that balances exploration and exploitation to a myopic algorithm that only maximizes the quality of the next recommendation. Our analysis shows that, compared with a monopoly, firms competing for users’ attention focus more on exploitation than exploration. When users are impatient, competition decreases the return from developing a forward-looking algorithm. In contrast, development of a forward-looking algorithm may hurt users under monopoly but always benefits users under competition. Competing firms’ decisions to invest in a forward-looking algorithm can create a prisoner’s dilemma. Our results have implications for artificial intelligence adoption and for policy makers on the effect of market power on innovation and consumer welfare. This paper was accepted by Dmitri Kuksov, marketing. Supplemental Material: The online appendix is available at https://doi.org/10.1287/mnsc.2023.4722 .

  PublisherDOI

• Supplementary Table 1 from Development of Autoantibody Signatures as Biomarkers for Early Detection of Colorectal Carcinoma

  2023-03-31

  supplementary-materialsOpen access

  <p>PDF file - 56K</p>

  PublisherDOI

• AI and Machine Learning

  World Scientific-Now Publishers series in business · 2023-06-19 · 1 citations

  book-chapterSenior author

  The following sections are included:IntroductionEarly AI — Expert Systems and ANNMachine LearningFuture ResearchReferences

  PublisherDOI

• Supplementary Table 1 from Development of Autoantibody Signatures as Biomarkers for Early Detection of Colorectal Carcinoma

  2023-03-31

  supplementary-materialsOpen access

  <p>PDF file - 56K</p>

  PublisherOA PDFDOI

Frequent coauthors

• Baohong Sun

  5 shared

• Young‐Hoon Park

  Cornell University

  5 shared

• Alan L. Montgomery

  Carnegie Mellon University

  4 shared

• Zhigang Lu

  University of Electronic Science and Technology of China

  4 shared

• Kinshuk Jerath

  4 shared

• Michael D. Smith

  3 shared

• S. Sriram

  Ross School

  3 shared

• Michael Trusov

  3 shared

Awards & honors

• Finalist, Marketing Science Institute/H. Paul Root Award
• Finalist, John D.C. Little Award
• MSI Recognizes Smith’s Liye Ma, Bobby Zhou for ‘Significant…