Distributed-Collaborative-Filtering-Book-Recommendation-System

It uses Dask as a Distributed Framework with Website Application using Streamlit. Inspired by the work of https://github.com/entbappy/ML-Based-Book-Recommender-System

Collaborative Recommendation System

• Collaborative filtering systems rely on user-item interactions.

• Users with similar ratings form clusters, facilitating the recommendation process.

• When recommending books, the system employs a cluster-based mechanism.

• The system considers either ratings or comments as its sole parameter.

• In essence, collaborative filtering assumes that if one user likes item A and another user likes both item A and another item, B, the first user may also be interested in item B.

• Challenges include:

  • The computational expense of managing a user-item nXn matrix.

  • Preferential recommendation for only popular items.

  • Potential neglect of recommending new items.

Data Used

We used the data from Kaggle that contains the Book names, User Ids and their Ratings.
Link to data: https://www.kaggle.com/datasets/arashnic/book-recommendation-dataset?resource=download&select=Ratings.csv

Algorithm Used

In training the Model, we used the KNN algorithm to cluster their ratings and in finding the suitable books to be recommended.

1. Load the dataset.

2. Set the value of k.

3. Iterate through the total number of training data points to obtain the predicted class.

4. Compute the Euclidean distance between the test data and each row of the training data, as it is a widely used distance metric.

5. Arrange the calculated distances in ascending order.

6. Extract the top k rows from the sorted array.

Distributed Framework - Dask

Dask is a parallel computing library designed to seamlessly scale and handle larger-than-memory computations in a distributed environment.

1. Convert the Pandas DataFrame to a Dask DataFrame

2. Find the index of the target book in a distributed manner

3. Compute the distances and suggestions in a distributed manner

4. Schedule the computation and gather results

5. Append the Book list into the array

Usage

Clone the Repository

git clone https://github.com/D3struf/Distributed-Collaborative-Filtering-Book-Recommendation-System.git

Open Anaconda Command Prompt and Create a conda environment inside the repository's directory

conda create -n books python=3.7.10 -y

conda activate books

Install the requirements

pip install -r requirements.txt

Now run the app.py

streamlit run app.py

Collaborators

• Jeanne May Carolino
• Almira Jill Garcia
• John Paul Monter --ME
• Aaron Paul Sicio