Case Study: University of Toronto advances large-scale language research with Google Cloud Platform

University of Toronto Using Vertex AI to reveal the words that make online reviews most helpful

The University of Toronto, a leading research and teaching institution in Canada, needed more computational power to study how language affects behavior in real-world settings. Working with Google Cloud Platform, Professor Sam Maglio and his team used cloud tools to analyze large-scale Amazon review data and test whether verb tense and wording influence persuasion and helpfulness.

Google Cloud Platform supported the research with Natural Language AI, BigQuery, and Vertex AI Notebooks to process nearly 250,000 words across 2 million Amazon reviews. The analysis showed that present-tense reviews were more helpful and persuasive than past- or future-tense reviews, and a related charity experiment found that present-tense wording led to the most donations.

University of Toronto

Sam Maglio

Professor of Marketing and Psychology