Skip to content

Neyung/UEH500

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Book Recommender using CF and GA

A Recommender System using Collaborative Filtering and Genetic Algorithm. We implemented a paper with a novel approach in building a robust recommender system named BLIGA to solve problems commonly seen in such systems, such as cold-start and sparsity.

The BLIGA consists of 2 main filtering levels:

  1. Semantic similarity. This is the level of similarity between features of items in a recommended list.
  2. Total predicted score. This is the fitness value of each recommended list, calculated by using the similarity between users and the AU (active user).

We modified the original BLIGA and tuned the hyperparameters. We also made some experiments to test the performance of the system.

Then we implemented a hyperparameter combination with top performance in a UI representing a book recommendation system in a digital library. The UI design is based on that of UEH Smart Digital Library, a key service of UEH University in Vietnam.

We have reported our process of designing, implementing, and testing the system and related problems in our final report. (Note: The content of the report is written entirely in Vietnamese)

Some related concepts

  1. User to Product Filtering

image

  1. Content-Based Filtering

image

  1. Hybrid Filtering

image

  1. BLIGA

image

DEMO

To be able to use the interface, please use: main.py - [Here]


MainWindow.2024-01-03.15-50-53.mp4

GUIDELINES

Our final Report - [Here]

BLIGA

The UI for BLIGA

Statistical Analysis of datasets

Releases

No releases published

Packages

No packages published