Introduction

This repository contains code for our paper Towards Job-Transition-Tag Graph for a Better Job Title Representation Learning accepted by Findings of NAACL 2022. Our implementation is based on the DGL package:

Requirements

Python packages required:

• Python 3.7
• PyTorch 1.11.0
• dgl
• scikit-learn

Dataset

CareerBuilder12 (CB12)

• Raw data downloaded from (https://www.kaggle.com/competitions/job-recommendation/data)
• Raw job transition data obtained from user_history.tsv
• Job title label assignment used O*Net-SOC AutoCoder

Steps of generating graph

• Generating job transition data: An example is given in 1.Prepare/1.Prepare_transition_cb12.ipynb
• Preprocessing job title and generating tags: An example is given in 1.Prepare/2.Preprocessing_cb12.ipynb
• Prepare graph files for homogeneous or heterogeneous graph: An example is given in 1.Prepare/3.Create_graph_*.ipynb

Job classification

• Examples are given in 2.Node_classification

Reference

Please cite our paper if you use our data or code:

@inproceedings{zhu-hudelot-2022-towards,
    title = "Towards Job-Transition-Tag Graph for a Better Job Title Representation Learning",
    author = "Zhu, Jun  and Hudelot, Celine",
    booktitle = "Findings of the Association for Computational Linguistics: NAACL 2022",
    month = jul,
    year = "2022",
    address = "Seattle, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2022.findings-naacl.164",
    doi = "10.18653/v1/2022.findings-naacl.164",
    pages = "2133--2140"
}