End State
- Limit screen time
- App locks
- Survey about child mentality and market research
- Content monitoring
- Simple message and alert the parent as per the words
- Pattern identification using ML models in social media context
- Create Mental health charts dashboard
- Parent community feed
- Firewalls
- Location Tracking (geofencing)
Features Details
-
Survey about child mentality and market analysis - DONE
-
Content monitoring - DONE
-
Simple Screen Reading of messages ad running in background - Facebook, Whatsapp, Instagram, Phone browsers, Gmail
-
Pattern identification and making sense of the context
-
Need to think on the bare minimum
-
Using ML models in social media context
-
Data collection - will be automatic through screen reading
-
Annotations - based on 3 categories
- Hate-speech - Hate-inducing, abusive, benign
- Discrimination - Gender, Racial, Ethnicity, Physical Appearance
- Mood - Depressing, Fun, Neutral, Romantic
-
Preprocessing
-
Classification Model
-
Reports for mood for the day
-
Mental health charts dashboard and Create Basic level 2-3 charts!
-
Location Tracking (geofencing) - Ability to live track location whenever necessary.
Later Phase
- Parent community feed
- Firewall- Tracking the websites
- Secure Location markings & Secure mode for the application ask the parent to punch the home location/ secure location, and when the kid is out of this secure areas, can have the app running in a secure mode which can help kids use panic buttons, other features can be thought of
- Designing the product that can be extendible in other local languages in future.
- Configurable level of abstraction.