A restaurant recommendation system for Bangalore.
Input a restaurant that you like from the list of 3649 restaurants available and then watch the model give recommendations for similar restaurants that you might like based on your input.
It is a simple algorithmic model that is based on certain handcrafted features like distance proximity, cuisines dishes offered, approximate price for two and more. For every pair of restaurants, it computes a score that is based on the weighted average of the handcrafted features and is then stored on a recommendations dictionary.
To generate recommendations, Depth First Limited Search is applied on the recommendations dictionary.
- distance_score - Computes the distance between the two restaurants.
- online_order_score - Whether the restaurants offer online orders
- booking_table_score - Whether the restaurants offer the facility of booking table
- rating_score - How similar are their ratings
- votes_score - How many people have registered their reviews
- type_of_restaurant_score - How similar are the restaurant types(like cafe,casual dining)
- dishes_score - How similar are the dishes offered
- cuisines_score - How similar are the cuisines offered
- approx_cost_score - How close are the approximate price for two person
- reviews_score - How similar are their reviews
Dataset - Zomato Bangalore Dataset
I am constantly working on making more refined and intelligent recommendations. Incase of any suggestions/feedback/issues feel free to drop me a mail.