The problem नारीSutram solves

Women are the backbone of our nation. The status of women depicts the social, economic, and mental condition in a nation, as stated by Jawaharlal Nehru. But the problem arises when the safety of our nation's women is bamboozled by a group of criminals, kidnappers, and rapists, moving freely around the streets. Unsafe streets restrict women's freedom & independence in life, causing them to feel underconfident, molested, and verbally taunted. While the Ministry of Women and Child Development has made several steps, including installing CCTV monitors at railroad stations with round-the-clock surveillance, outfitting phones with emergency helplines (181) and panic buttons, etc. नारीSutram is a platform that functions as a personal safety guide for women, is our contribution to avoiding this dreaded problem in India.

1)Detecting the safest route:

Upon entering the Source/ Current location of our user & the Destined location, नारीSutram predicts the safest road for our user to traverse amidst all the other paths.

2)Detecting the safety score of the current position:

Along with the Travel distance and the approximate time taken, our platform gives a Safety Score based on the previous surveys and datasets by people who found the road free from miscreants.

3. Automated SOS Button:

Upon clicking the SOS button, the live location of our clients will be sent to their trusted contacts list, with a read alert that she is in danger.

4)Comprehensive Feedback form:

We used feedback form to get few details like lighting that are not originally extracted through public datasets. User is asked to enter a rating and remark, sentiment analysis is done on the remark and a rating between 1-5 (1 suggesting unsafe and 5 suggesting safe) is assigned to it,an average of this score and the rating entered by user is taken, and if it's above a certain threshold (which we have taken to be 3) then the route is considered to be safe by user.

Challenges we ran into

1. The Machine Learning Model we used, K-nearest neighbour model has some intrinsic errors while finding the approximate value of 'k'. So, in the initial stage, it was an exhaustive task for us.

2. The integration of our trained model with the maps was a hectic task, we had to do comprehensive research on it before the integration took place.

3. Finding the Safety of the other paths and comparing it with our proposed best roads was also a challenge for us initially.