- Stress can leads to depression / hypertension and few cases student may take drastic steps. In essence, stress is not good, and caregivers need a tool to understand psychological and emotional state of students so that they can provide timely support
- Because we believe that helping people is the right thing to do :)
- Six people team up to develop a technology that enable early detection of stress
- A website that helps caregivers detect depression / suicidal feelings of students in using social media
- Before 27/09/2019, finish a website for caregivers, with a dashboard so Caregivers (Teachers, Counsellors & Parents) are alerted and Students can get timely help
- We succeed when our app can do the following tasks:
- Evaluate students' emotional states every 5 minutes
- Send alert to the caregiver within 15 min when time-critical action is required
- Language: Python
- Jupyter notebook: to develop model
- ML library: Pytorch / TF
- Pre-trained word vectors: GloVe / Google Vectors
- Language: Python
- Jupyter notebook: to develop model
- ML library: Pytorch / TF
- Language: Python / JS
- Framework: Django / Flask
- Tweet Scraping library: TBD
*Facebook chosen for this project as it is the most popular social media
- Sunday - Friday: develop
- Saturday: close tasks, give feedback, define new tasks
- WhatsApp for chats
- GitLab for version control
- In the main dashboard go to Issues -> New Issue
- Name the sub-issue using the following syntax: [name_of_parent_issue] name_of_sub_issue
- Tip: adding due-date may help you keep track your progress more efficiently.
- NL Engine:
- Kane
- Adithya
- IP Engine:
- Mayur
- Dhananjay
- Web App:
- Minal
- Giang
- [1] Separate components (01/09/019 – 07/09/2019): Develop engines, webapp separately
- [2] Fist version (08/09/019 – 14/09/2019): Integrate engines with webapp
- [3] Second version (15/09/019 – 21/09/2019): Improve performance of application components
- [4] Final Version (22/08/019 – 27/09/2019): Develop additional functionalities