The problem wayfer solves

Background: On any normal day, while we're out travelling in any city across the world, we often find ourselves dependent on HERE Maps. We use it to find the most convenientpath to our destination. It offers us the shortcuts and less traffic prone area to reduce our travel time. But, what it fails to inform us is, how safe is the path that we travel on. In the process of finding the optimum path, Google Maps often leads us to roads that may be part of the route, but are extremely deserted at certain times of a day, and often quite unsafe. The situation has been welldescribed in an article of Financial Times as follows:- "So why doesn’t the app provide safer walking directions to women, or indeed anybody else, at night? It knows where we are, it knows what time the sun sets. It knows the population density of the area, and crime statistics are geotagged and publicly available in most major cities. It can even see which areas people walk through with impunity during the day but avoid in the dark." What we are doing:-
We divided the target city into 150 -300 head nodes, and made a graph of the city.Based on past data and current data of a road we Update hourly weather the road is safe or not.Not only that we We also continously scrape twitter data to seacrh for any accidents which may occur , and show them onto our map.We made a chat application which allows users to ask queries only to nearby peoples,(implemented using graph(people residing in the neighbouring nodes)).As this is a community based App Users can mark their house as safe house.If any woman is attacked, she can press the panic button, alerts will be sent
to nearby people(implemented using graph(people residing in the neighbouring nodes)), and family members.We are not only showing a road safe/unsafe based on our model(which takes into consideration jam factor confidence and past data), but also add data added by real users to avoid false positives and true negetives. Lastly family can track them as wel

Challenges I ran into

Creating a Graph of a city was really a challenging task. It was the most difficult part.Not only that but Also diviind the areas amongst the head nodes as also a daunting task.
Scrapping twiiter was simple But extracting Names was a difficult task.We used NER(named entity recognition) and also specified some general rules to extract location name, and used reverse geocoding API of HERE to get there coordinates and show their location.Creating the back end server of this project was also very challenging especially the chat application part as we needed to switch node based on geolocation and render the messages only to nearby peoples.Location tracking part was handled using pubnub.
We used a simple Machine learning Model for predicting weather a road is safe or not(decision tree to be precise).