Sharing Part of my project work done while taking Specialist Diploma In Applied Artificial Intelligence
A NLP based restaurant recommendation system using Dialogflow and deployed on Telegram interface.
Abstract For This Project YELP is a business directory service and a crowd sourced review forum. In this project, I will focus on the reviews for restaurants in Las Vegas. The central idea is that the sentiment and semantic meaning of a user’s text input can reveal useful information about a user’s preference for a restaurant, and in consideration with other preexisting ratings for the restaurant such as average restaurant rating, other user ratings... etc. The aim was to produce a recommendation system that takes in this composite of existing ratings, fresh input from user’s sentiment, user’s likes and dislikes, together with a predicted user’s rating from Random Forest model produce a hybrid recommendation system to generate a list of recommendations. Both Cosine Similarity and Euclidean Distance are used to determine similarity. A Dialogflow agent is used to manage conversational flows delivered through TELEGRAM interface. This virtual assistant will be available on the internet 24/7.
