There are 2.2 billion children in the world and they constitute 28% of World’s population. The age range between 10 to 19 years make up 16% of World’s population COVID-19 pandemic and lockdown has led to short term as well as long term psychosocial and mental health implications for children and adolescents. A growing number of concerns related to emotional health and issues of these children such as anxiety, acute stress, loneliness, have accelerated due to lockdown.

Social media fosters connection and engagement in ways that can build wellbeing among young people. Young people are now sharing their online persona by expressing their thoughts, moods, emotions, and even their daily struggles with mental health on social media. During mental illness typically they post Self-portrait facial images and correlated text in social media such as Instagram, Twitter and flickr. Researchers worldwide have started parsing and analyzing social media data and made these into datasets that can be utilized to model mental health phenomena. We decided to contribute to this project by enabling the researchers/developers an API where they can feed their dataset and instantly get the negative emotion analysis and text sentiment analysis from major AI Cloud providers and present it to them for further analysis.

What it does

Mind Your Mind is a PostMan public workspace for researchers and developers to analyze the Negative Emotion Index of social media facial images through Face Emotion APIs and social media text messages through Text Sentiment APIs utilizing major AI Cloud providers - Azure Cloud, AWS Cloud & Google Cloud. A Slack notification is also sent as an alert indicating the negative emotion index and negative sentiment index of the young individual.

The App provides the following workflow:

  • The facial images and the associated text derived from social media platforms can be derived from the researchers’ data set (or) directly parsed from the social media platform

  • Social Media Facial pictures are passed through Facial Emotion Identification API provided by major AI Cloud providers (AWS, Azure and Google cloud) and computation is done to get a standardized Facial negative emotion index (0 to 10 – 0 is low while 10 is high)

  • Social Media Text messages are passed through Text Sentimental Analysis API provided by top 3 AI Cloud providers and computation is done to get a standardized Text negative sentiment index (0 to 10 – 0 is low while 10 is high)

  • Based on the Facial Negative Emotion Index and Text Negative Sentiment index, a high alert message is provided to the human coordinator for the immediate intervention that needs to be done for the mentally ill person

How we built it

  • Refering to Postman Learning Center and other resources provided in Resources and Inspiration pages

Challenges we ran into

  • Getting real world public data sets of mentally ill young adolescents was a challenge and privacy policy is strict for these data sets

  • Different AI cloud providers have different standardizations for emotion analysis and sentiment analysis, consolidating them was a bit of challenge

Accomplishments that we're proud of

  • Implemented the Postman API that even low code developers would be able to understand and take advantage and further build and evolve the solution

  • Utilized the Postman API in such a way that it not only benefits the researchers/developers but aids in the social good for kids especially affected during this Covid-19 pandemic

What we learned

  • Learned about the API testing platforms and advantages of Postman API in comparison to others

  • Learned about the new approach of parsing the social media data by researchers for the mental well-being of young ones

What's next for Mind YOur Mind - Mental Well-being of Kids during Covid-19

  • Build an API that utilizes the multi-modal approach in getting voice tonal analysis, analysis of neural co-ordinates with wearable brain sensors, analysis of body physiological parameters with smart wearables so that we can get a comprehensive mental well-being of kids

  • Evolve the API by sharing it with researchers and developers so that a comprehensive and accurate multi-modal approach maybe followed

Built With

Share this project: