Can machine learning help classify natural health products and describe their use?

Project Summary

Natural health products (NHP) encompass a variety of substances, including vitamins, probiotics, and homeopathic remedies. They are widely used in Canada, notably by women, aged individuals, and those who have a chronic disease such as diabetes. However, the potential benefits of NHP are not clearly established, in part due to the lack of an official classification system, which makes comparison and replication across clinical studies challenging. While rich information exists on the web, it is non-exhaustive and found as plain text. We will use machine learning to extract this information from many sources and organize it into a classification system. Then, we will use this classification to describe NHP use in the CLSA cohort, including estimated costs, and evaluate its impact on cardiovascular diseases, diabetes, and cognitive decline.