Restaurant Recommender Systemon Machine Learning with Graphs

InspirationImagine a world where choosing the perfect restaurant is as simple as a tap on your phone, with recommendations tailored precisely to your tastes and preferences. Our project is inspired by this vision, aiming to revolutionize how we discover dining experiences. By harnessing the power of machine learning and graph representation learning, we aspire to create a recommender system that not only understands the intricate relationships between users and restaurants but also predicts ratings with remarkable accuracy. This innovation will transform everyday dining decisions into personalized, delightful adventures, making the process of finding great food as enjoyable as the meal itself.

What it does

My project, the Restaurant Recommender System, predicts users' ratings for restaurants by analyzing user preferences and restaurant characteristics. This allows for personalized recommendations, enhancing the dining experience.

How we built it:

Developed the system using two approaches: feature engineering and graph representation learning. Feature engineering involves extracting relevant characteristics of restaurants and user preferences, while graph representation learning captures the structure of the user-restaurant interaction graph into embedding vectors. We then used these inputs in a feed-forward neural network to predict ratings

Challenges we ran into

We faced several challenges, including selecting the most relevant features for the model, ensuring the graph representation accurately captured the complexity of user-restaurant relationships and balancing the computational efficiency with prediction accuracy.

Accomplishments that we're proud of

We successfully demonstrated that machine learning models could be effectively used with graph data. Additionally, we showed that graph embeddings could achieve similar or better prediction results than hand-crafted feature vectors, marking a significant advancement in recommendation systems.

What we learned

We gained valuable insights into the intricacies of feature engineering and graph representation learning. We also deepened our understanding of the challenges involved in creating accurate and efficient recommender systems.

What's next for Restaurant Recommender System on Machine Learning with Graphs

Next, we plan to enhance the model's accuracy by incorporating more sophisticated graph neural network techniques and expanding the dataset to include more diverse user and restaurant attributes. We also aim to deploy the system in real-world applications to further validate its effectiveness and improve its performance.

Built With

• django
• matplotlib
• numpy
• pandas
• postgresql
• python
• rest