Project Sparrow
Project Sparrow is one of ROEB's Transformation initiative which started in April 2021. The project aims to create a digital library of online data enabled by web scraping technologies of emerging health products, techniques, technology, processes, trends etc, for the purpose of informing Compliance &Enforcement decisions. We are currently in a research phase - defining the problem statement, discovering what other partners have done in this space. The project anticipates taking my pivots as new information is discovered.
Project Context
In fulfilling its mandate, ROEB applies a sound and robust risk management approach to decision-making that prioritizes the overall health and safety of Canadians. Risk management tools and Frameworks are further complemented by the data which is available, useful and of good quality. ROEB primarily uses internal collected data to inform its C&E decisions, which is primarily used to:
- Respond to media requests
- Equip Canadians to make informed decisions about their health
- Communicate clear expectations for regulated parties
- Identify areas for improvement
- Inform operational C&E decision-making through identification of trends, emerging issues, and areas of focus
- Measure performance
One of the many challenges ROEB faces when applying risk management framework is the availability of data. For instance for inspections and compliance verification, we rely on inspectors comments and industry completed forms to collect data to inform our decisions. This data is often difficult to work with as it may not meet one or more of the core data properties:
- Relevance and usefulness
- Accuracy
- Timeliness
- Accessibility and clarity
- Comparability
- Consistency
- Completeness
The question becomes: is there a way to further complement the data that ROEB collects to further inform decision making? There are numerous success stories in industry that capitalize on publicly available data to inform business decisions. ROEB would like to explore the use of public source of data to further refine, inform, or validate current C&E decisions.
The Waterloo Experience (WE) Story
An opportunity was presented to Health Canada to be part of the Waterloo Experience - a program which provides students in-demand Artificial Intelligence (AI) and business skill development (co-created with AltaML, Microsoft and Industry/Public partners), team-based project experience, and real-world industry exposure. Health Canada is one of the partners that will bring a business problems to these students so they can apply AI and ML to solve them.
The Problem: Site Risk Profiling. Decisions informed with data that is driven by risk is a prime focus for all programs in ROEB. Health Canada is proposing to use Site Risk Profiling as the use case for this opportunity. Existing approaches to Site Risk Profiling is being conducted with HC data applied with multi-variant statistical analysis. It is unknown, however, if this approach has biases or gaps that can create risk to the compliance and enforcement mandate. To validate Health Canada’s approach to Site Risk Profiling, we are proposing to leverage public data via Twitter to assess Machine Learning potential and compare our approach to a solution provided by WE Accelerate.
Repository
Click here to access the Sparrow files.
