SAP Brazil and Amazon Web Services join Africa agency to improve the “Algorithm of Life”

  • Tool can identify profiles of people with emotional problems based on text content whose language signals depression 
  • SAP Brazil and Amazon Web Services will collaborate to improve the algorithm and allow the project to scale 

SAP Brazil and Amazon Web Services (AWS) join the team of the Brazilian advertising agency Africa agency, to improve and scale the “Algorithm of Life,” a tool developed to prevent suicide by analyzing public posts on Twitter. With the support of leading technology companies, the tool will be enhanced to increase its effectiveness and reach.

The SAP partnership will allow the solution and database to be replicated in the SAP HANA Cloud environment for algorithm refinement by SAP experts, such as database enhancements and predictive intelligence (SAP Analytics Cloud). The tool will be able to identify behavioral changes in monitored profiles and signal when immediate actions are required. Together, SAP and AWS will provide a team of data analysts and solution engineers who will work to continually improve the “Algorithm of Life.”

Luciana Coen, SAP Brazil Communication and Corporate Social Responsibility (CSR) director, believes that initiatives like this have a strong impact on the identification of emotional vulnerabilities and people’s access to support services. “Technology has the ability to scale in order to reach more people. The Algorithm of Life is a project we have embraced because we strongly believe it will make a significant difference in the lives of many”, she said.

The Algorithm of Life platform will run on an Amazon Web Services cloud environment. AWS will also designate professionals from its solution architects team, to work on the application development. The original solution developed by Africa agency and Bizsys, will go through the Well-Architected program’s analysis in order to improve the operational excellence, safety, reliability, performance and cost optimization.

“We are a company that is obsessed with customers, and putting our entire cloud innovation history at the service of people’s welfare, no matter wherever they are, is a way of giving a higher sense to our first principle of leadership,” said Cleber Morais, AWS general manager for Brazil.

In addition to the technical expertise provided by SAP and AWS, Africa agency considers that the partnership will add an data-based analytical view to ensure full control of the tool. It will also provide a real-time diagnosis of the monitored profiles that can also support activities that engage society in projects to clarify and combat depression as well as to provide material for public health actions.

“The Algorithm of Life was born from the Africa agency’s idea that now reaches a new scale. This has always been one of our goals since the project’s beginning. So it is gratifying to put this new partnership into practice and see the communication and technology areas joining forces to generate more and more positive impacts,” said Sergio Gordilho, Africa co-president and CCO.

How does the Algorithm of Life work?

It is estimated that over a million people commit suicide each year worldwide.  This represents one death every 40 seconds. In Brazil, this scenario is also present. Every day, 32 Brazilians deliberately take their lives, with 11,000 lives lost in 2018.  According to the Ministry of Health, suicide by young people aged 15-19 has increased by 20% in the last five years.  Suicide is currently the fourth most frequent cause of death of youth in Brazil.

Studies by mental health experts* indicate that people suffering from depression recurrently use a particular group of words, as a kind of “depression grammar”, to indicate the disease symptoms even at an early stage. Obviously, these signals are also present in behaviors on social media.

The Algorithm of Life was created using this data to identify a huge variety of words and sentence that may point to symptoms of depression in users’ public posts on Twitter. After the first phase for word and expression identification, a careful check is carried out by a trained team to consider context, ironies, recurrence of terms and periodicity.

Watch the video for more details about the project on https://www.youtube.com/watch?v=tmkESsucvHY

Sources:

Prevent suicide: a global imperative, OMS – https://bit.ly/2t9xMYu

Nature Science magazine’s article – https://www.nature.com/articles/d41586-017-08307-0

*References:

Predicting Risk of Suicide Attempts Over Time Through Machine Learning, Colin G. Walsh, Jessica D. Ribeiro and Joseph C. Franklin, Association of Psychological Science of Florida State University. https://bit.ly/2DwOYgM

  • In an absolute state: elevated use of  absolutist words is a marker specific to  anxiety, depression and suicidal ideation, dos autores Mohammed Al-Mosaiwi and Tom Johnstone Department of Psychology, School of Psychology and Clinical Languages, University of Reading. https://bit.ly/2UGhySu