Federal health researchers are looking to mine social media to more quickly identify suicide-related behaviours, instead of relying on woefully outdated data.

What the Public Health Agency of Canada wants is an artificial intelligence program that combs social media platforms for keywords to give its researchers a better view of trends and risk factors linked to suicide-related behaviours.

The proposed pilot project, outlined in federal bidding documents, would also give a window into suicide risks based on sex, age, ethnicity, and geographic location.

Story continues below advertisement

The agency says it is not, however, looking to make the pilot project a tool to predict suicides or identify specific people at risk of attempting to die by suicide.

Suicide is the second-leading cause of death for Canadian teens.

The health agency produces regular reports on suicide trends using death and hospital data, but the information can be up to five years old when the agency gets it.

The agency says to better prevent suicide, it needs more up-to-date information of “suicide-related behaviours, which occur further upstream,” to figure out how to craft earlier interventions for those who need the help.

The bid documents say social media sweeps would start in June 2019. By the end of next year, the agency expects to produce monthly social media monitoring reports.

The cost to create a program to sweep sites like Facebook and Twitter for information is expected to be up to $150,000, excluding taxes, according to the request for proposals.

The contract runs until March 2020.