Research

Data, AI, and Consumer Insights

My research explores how businesses and consumers interact with digital content, particularly in e-commerce and online platforms. I analyze large-scale datasets, spanning product descriptions, consumer discussions, and metadata, to extract insights that enhance digital marketing and business decision-making. While much of my work focuses on text analysis, I also explore interdisciplinary approaches that incorporate visual and contextual data.

I investigate questions such as:

How can AI improve the accuracy of product information in large datasets?
How do consumer discussions shape product perception and brand engagement?
What role does unstructured data, text, images, and metadata play in digital marketing analytics?

Using NLP, machine learning, and generative AI models, I analyze patterns in consumer-generated data to uncover insights that drive better marketing and business strategies.

Ongoing Projects

I am currently working on projects related to:

Sensory Cues and Consumer Perception in Digital Environments
I explore how verbal and visual elements, especially color, shape consumer attention, expectations, and trust in online shopping contexts. This stream combines large-scale review analysis, NLP, and generative AI models to study how consumers process color-related cues in text and images.

AI-Powered Product Data Enhancement
Using LLMs and lexicon-based tools, I develop techniques to improve product metadata quality (e.g., attributes like color, material) across large e-commerce datasets.

Consumer Conversations & Market Trends
I apply topic modeling and rhetorical analysis to understand how consumer-generated text reflects product experience and influences engagement metrics like review helpfulness.

Color Impact 2025 Conference (CI25)

I presented my research at the Color Impact 2025 Conference (CI25) in Rochester, NY, showcasing methods for extracting product color attributes using AI and lexicons. The talk highlighted tradeoffs between accuracy, coverage, and automation in processing large-scale Amazon metadata.

Selected Projects and Working Papers

How to Persuade People to Get Vaccinated: Be Emotional or Rational?
With Milad Hour, Golshad Parsimoghadam, and Ali Jamei
This project uses rhetorical theory and a dataset of over 13,000 tweets to explore how emotional (pathos), logical (logos), and credibility-based (ethos) appeals affect public attitudes toward vaccine messaging. We find that emotional appeals significantly outperform logic-based messaging in generating engagement, with implications for health marketing and communication strategy.

Diversifying AI Tool Use in Undergraduate Business Classrooms (Position Paper)
With Golshad Parsimoghadam
This teaching-focused project examines how business students interact with AI tools during marketing coursework. Findings reveal overreliance on familiar tools (e.g., ChatGPT) and limited experimentation with newer or more specialized platforms. We propose structured interventions to promote AI literacy through comparison-based assignments and peer learning.