Publication Venue Recommender System

This repository contains code for the project Modular-Hierarchical Attention Based Scholarly Venue Recommender System

The model is built using a hierarchical attention based Bi-LSTM. The dataset used is A-Miner Dataset which contains approximately 18Lac Papers related to Computer Science distributed accross 1.5Lac Journals. We have taken only those Journals which have published more than 500 papers.

Requirements

• Anaconda3
• python3
• PyTorch version 1.7.1 (Installation instructions)
• Scikit-learn (run pip install scikit-learn)
• Install CUDA 10.1 if you need to run the code on a GPU.

Usage

Convert your dataset into the format mentioned in format.txt
Extract data from format.txt type files into pickle dumps using extract_data.py
Generate the PyTorch compatible dataset using the text and authors vocabularies using generate_dataset.py
To train a model on the training data, use train.py
To test the model on some test data, use test.py

The code is meant to run with both CPU and GPU support.