Skip to content
Branch: master
Go to file

Latest commit


Failed to load latest commit information.
Latest commit message
Commit time


Recommender systems are now popular both commercially and in the research community, where many algorithms have been suggested for providing recommendations. These algorithms typically perform differently in various domains and tasks. In this repository we deal with different algorithms we used to build a recommendation system for suggesting how likely a user would answer a question using the byte cup data.

General instructions

All the code is in python and can be easily run by just cloning/downloading the repository. It already has all the features generted and the code for executing different algorithms can be found in the Expert-recommendation-system/code directory.

Running the code

You need to have python 2.7 setup on your machine for running code on this repository. The code could be simply run with the following command python [file_name]

Instructions for Running various Algorithms

  • Linear and Logistic regression
    Files associated :-,

  • Collaborative Filtering

    • User-based Collaborative Filtering
      Files associated :-,

    • Item-based Collaborative Filtering
      Files associated :-,

  • Content-Based Method

    • Content-Boosted Collaborative Filtering
      Files associated :-,,,,,

    • Content-Based Method With K Nearest Neighbors(KNN)
      Files associated :-,

  • Hybrid Method
    Files associated :-

  • Neural Networks
    Required Libraries :- TensorFlow Files associated :- /code/neural net/ and all the assoicated files are present

  • XGBoost
    Files associated :-,,,,,

  • Sparse Linear Method (SLIM)
    Files associated :-,

  • SVD++
    This was run using various libraries like MyMediaLIte and librec using some of the feature files directly.

  • MyMediaLite
    Files associated :- This is a library and the commands to run the same are in the /MyMediaLite-3.10/nb.txt

  • Matrix Factorization
    Files associated :- MatrixFactorization.ipynb this is an iPython notebook you need ipython to run this.


Recommends an expert for a set of domains, trained over a prior data.




No releases published
You can’t perform that action at this time.