The problem Militia solves

With India topping the charts in terms of threats to national security, how can a civilian assist the military? Ironically, the greatest threat to national security arises when individuals take a step back, fearful of the unforeseeable consequences thereby hindering its prevention. A collaborative action taken from both sides is the sole way out to prevent any potential attacks.
But, where to address crucial tips in a chaotic situation that might assist the military in their operations?
Even with information at hand, it is difficult for the military to distinguish a legit tip from another spiteful suggestion.
To address all these issues, we present to you our project: Militia.
The motive behind our project is to provide an interface for civilians to collaborate with the military. The interface comprises a form where civilians can report any suspicious activity anonymously. The data sent from the form is encrypted and can only be accessed by a single admin, ensuring security. Different prioritization techniques are implemented to ensure the authenticity of the tips.

In order to inculcate authenticity, our portal requires the users to register themselves with basic information including Aadhar verification by Verhoeff algorithm and mobile number regex validation. Authentic users can log in, impart tips, and read notifications. The future idea behind registration is to send push notifications in time of evacuation to them.

To ensure the security of the information being reported, we have encrypted the message using Django Cryptography along with Fernet Encryption which can only be accessed and decrypted in the admin panel by another password protection while decryption.

Prioritization of tips is evident problem that is addressed by implementing Filters in the admin panel. Furthermore, a LSTM Machine Learning Model trained on a dataset of 3000 for spam classification aims to shift the focus on the issues with high priority.

Challenges we ran into

Decryption:
Ensuring no loopholes of breaking into the admin panel, we aimed to not decrypt the message to the admin without further password verification. Thus, decryption of data at the admin panel was among the main challenges faced in this project which was later resolved.

Deployment:
Bounded by time constraints, finding the correct dataset was indeed a problem. Though we were able to train the spam classification model, integrating that with Django and deploying the website is still a paramount challenge to overcome.