Our system uses Twitter and Google Trends data to track indicators of Physical Activity, Sedentary behaviour, and Sleep quality (PASS) of Canadians.

Big data from the Internet tells us a lot about what people are interested in. So why not use it to see if tweets and Google search data can give us any insights into how much Canadians care about physical activity, sedentary behaviour, and sleep quality.

Since November 2017, we have been collecting location-tagged tweets and Google Trends data to see how well these unconventional data sources can be used to predict the prevalences of common diseases, such as: diabetes, obesity (body mass index), and mood disorder.

Please choose 1 of 3 behaviours and a week to be analyzed below:

Twitter PASS Indicators Google Trends PASS Indicators
Canadian Community Health Survey Statistics :
Following are the top 5 locations for physical activity
Twitter PASS Indicators

Twitter is a social networking site with up to 190 million users worldwide producing up to 50 million tweets daily. Twitter is utilized as a micro-blogging site to share opinions, mood, and activities, including health-related conditions and behaviours.

Our Twitter PASS indicator tells us what portions of the tweets collected are relevant to the Physical Activity, Sedentary behaviour, and Sleep quality in the first person point of view. To do this, we used a machine learning model that was trained with a set of example tweets (hand-labelled by our team) to classify whether a tweet is relevant in the first person or not. The tweets are then aggregated by location and mapped according to the Canadian Health Regions. Only Health Regions with a sufficient number of tweets were mapped.

Twitter PASS Indicator for each Health Region across Canada

Twitter trends may change overtime. Let’s take a look at the historical data that we collected.

Twitter PASS Indicator for Canada Overtime

Google Trends PASS Indicators

By examining vast number of searches on the Internet, we are given a unique perspective on what Canadians are currently interested in and curious about.

The Google Trends Indicator represents the total interest of five selected topics chosen by our team for each of the three PASS components. Google Trends data provides us a random sample of all Google searches from Canada since 2004, and across some major metropolitan areas. The data is anonymized (no one is personally identified) and categorized (grouping similar queries into a topic). Google Trends data are provided in relative search volume (RSV).

Google Trends PASS Indicator for each Health Region across Canada

Google Trends PASS Indicator for Canada Overtime

Compare Two Locations

Pick an indicator and different locations in Canada to begin comparing:

Twitter and Google Trends data provides an interesting and powerful lens into what social media users are curious about, as well as their underlying health-related behaviours. We look forward to continuing the conversation.