Making Big Data More Inclusive
A newly appointed DLSPH Faculty member and data pioneer is on a mission to help public health practitioners hear the “quiet voices” not traditionally captured in online data gathering.
Neil Seeman has become a world-leading expert in uncovering the perspectives and experiences of people who don’t typically take surveys or post their opinions online. By developing alternatives and complementary approaches to Internet “scraping” algorithms and online panel surveys, he hopes to make Big Data a more accurate and useful tool in improving population health and health systems.
“We talk too much in Big Data about the volume of data, but in my opinion, it’s really about inclusivity,” says Seeman. “You want to have as diverse an audience as possible who are giving us insights. We’re trying to evolve away from WEIRD data (Western, Educated, Industrialized, Rich, and Democratic) because WEIRD data lends itself to bias and limits insights.”
Seeman founded and heads the global trend-tracking and predictive analytics firm RIWI Corp., which works with corporations, NGOs, and governments to collect novel survey and message testing data in all countries. He was just named a Senior Fellow at DLSPH’s Institute for Health Policy, Management and Evaluation (IHPME) and a senior academic advisor at the School’s Investigative Journalism Bureau, where his firm’s alternative survey approach was an important element in the ground-breaking Toronto Star series, Generation Distress. He has strong roots at U of T: His father was Philip Seeman, who first identified dopamine receptors as being the primary site of action of antipsychotic drugs. His mother is psychiatrist Mary V. Seeman whose research into schizophrenia has focused on the role of sex and gender. Both his parents are Officers in the Order of Canada. Seeman is currently writing a book about mental health and business culture, drawing on the “family business of dopamine”.
Seeman is particularly interested in exploring digital spaces to compile and learn rapidly about evolving barriers to the uptake of vaccines across the globe, effective mental health strategies and structures, and about the diverse challenges facing caregivers. At DLSPH, he’d like to advance his conceptual framework for an open-access system that roves the online world for real-time signs of emerging treatments or trouble – more digging than scraping. His beta version of this new conceptual approach was recognized in February as a Top 30 Finalist for the Sternfels Prize for Drug Safety Discoveries.
“My ambition for the global data ecosystem is ambitious – I want to help build a global, privacy-compliant healthcare grid of both misinformation, and of people’s self-reports on their chronic conditions, how they’re treating them, including off-label uses, and the reported side effects,” he says. “My hope is that people with the same suite of challenges in Bangalore, India may be able to learn from someone with the same suite of challenges in Toronto.”
Seeman is an “infodemiologist” – a researcher interested in how information and misinformation spread online can either help or harm public health. He invented and deployed a tool to track the spread of information online about H1N1 to gain insights on how people produce, perpetuate, and consume information relating to a pandemic. More than a decade ago, that tool operated in real time. Today, with COVID denialism and vaccine hesitancy sweeping social media networks, Seeman sees an urgent need for much bigger, more ambitious real-time data collection tools – ones that can pivot to any part of the world and quickly digest what patients and caregivers are saying at the regional, local, or even neighbourhood level.
“When guiding healthcare policy or understanding dynamic human behavior, I’ve learned it’s vital that we hear about the needs and fears of quiet voices – such as the homeless or unbanked or migrant populations, and the growing digital populations who do not answer traditional online surveys or who never participate in social media,” says Seeman. “To understand how to build effective public health systems at scale to reduce chronic disease, and nurture a data-driven culture of prevention, infodemiology can be global and local,” he added. “A machine-learning model that reports what quiet voices with different health profiles express in rural parts of China should be rapidly accessible to help policy-makers in Italy. At Dalla Lana, we can make that happen – with passion, collaboration, and sustainable world-wide impact.”