https://github.com/manisharora96/book-recommendation-systems

In this project i make a recommendation system for books if you want to read a book this recommendation system recommends some best book to you for reading.
https://github.com/manisharora96/book-recommendation-systems

Last synced: 1 day ago
JSON representation

In this project i make a recommendation system for books if you want to read a book this recommendation system recommends some best book to you for reading.

• Host: GitHub
• URL: https://github.com/manisharora96/book-recommendation-systems
• Owner: manisharora96
• Created: 2022-08-07T08:09:52.000Z (over 2 years ago)
• Default Branch: main
• Last Pushed: 2022-08-07T08:17:33.000Z (over 2 years ago)
• Last Synced: 2024-06-29T15:01:09.584Z (7 months ago)
• Language: Jupyter Notebook
• Size: 2.83 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Book-Recommendation-Systems

## Goal

The goal of this project is to make a recommendation system which will recommend the user a book, based on the search results given input by the users.

## Dataset

The dataset which is used here, is collected from Kaggle website. Here is the link of the dataset : https://www.kaggle.com/jealousleopard/goodreadsbooks. I have uploaded the same in [`Dataset`](https://github.com/abhisheks008/ML-ProjectKart/tree/patch-45/Book%20Recommendation%20Systems/Dataset) folder too, you can access that!

## What Have I done?

1. Importing all the required libraries. Check [`requirements.txt`](https://github.com/abhisheks008/ML-ProjectKart/blob/patch-45/Book%20Recommendation%20Systems/requirements.txt).

2. Upload the dataset and the Jupyter Notebook file.

3. Data Description

4. Data Processing

5. Data Visualization

6. Outliers Handling

7. Machine Learning Models

    - KNN

    - t-SNE

    - DBSCAN

8. Testing the models

9. Conclusion

**************************************

## Libraries used

|Numpy|Pandas|Matplotlib|Sklearn|Seaborn|XgBoost|isbnlib|

|-|-|-|-|-|-|-|

|progressbar|copy|scipy|mpl_toolkits|newspaper|tqdm|

************************************

## Conclusion

* **Clustering** is the best way to represent this project.

* we have deployed three different types of clusters, such as, **KNN**, **t-SNE** and **DBSCAN**.

* After deploying all the algorithms, we have seen that more or less all the models are showing **similar results**.

* Hence, to develop the recommendation system one can use **any of the three algorithms** used here.

* But, using **DBSCAN algorithm** provides the user better experience in terms user data given, so I will recommend to use DBSCAN algorithm over the KNN and t-SNE clustering.

* DBSCAN Model is having the accuracy of **92.56%**.

*******************************