Inspiration
We were all talking about times where the places that certain travel blogs recommended were far away from where we currently were and did not factor in factors such as what we liked to do or what the ideal weather was to visit that attraction.
What it does
The app recommends travel places based off of user preferences in activities, distance, and rating.
How we built it
We accessed the Places API to get the data about points of interest to train our model on seven cities in Italy. Using those trained weights, we could make new predictions for a new city.
Challenges we ran into
We ran into challenges adding the weather feature and trying to link the model to our front-end.
Accomplishments that we're proud of
We are proud that we got the model to work for some of the limited features.
What we learned
We learned about K-nearest neighbors and how we could use it to find close recommendations.
What's next for Tripology Assistant
We would definitely try and incorporate a few more features in the model such as weather or maybe even a personality type of the user.
Log in or sign up for Devpost to join the conversation.