/name-disambiguation

this is a pipeline of author name disambiguation.

Primary LanguagePython

name-disambiguation

this is a pipeline of author name disambiguation.

If this code helps you, please cite the following paper: Qiao, Ziyue, Yi Du, Yanjie Fu, Pengfei Wang, and Yuanchun Zhou. "Unsupervised Author Disambiguation using Heterogeneous Graph Convolutional Network Embedding." In 2019 IEEE International Conference on Big Data (Big Data), pp. 910-919. IEEE, 2019.

Basic requirements

  • python 3.6.5
  • networkx 1.11
  • gensim 3.4.0
  • sklearn 0.20.1
  • numpy 1.14.3
  • pandas 0.23.0
  • tensorflow 1.10.0

Note: you are recommended to run this pipeline on windows.

Run this pipeline

# step 1: preprocess the data
python data_processing.py

# step 2: train the GRU based encoder to learn deep semantic representations
python DRLgru.py 

# step 3: construct a PHNet and generate random walks
python walks.py

# step 4: weighted heterogeneous network embedding
python WHNE.py

# step 5: generate clustering results
python evaluator.py

Data

you are recommended to use the word2vec model we pre-trained to generate word embeddings of publication titles via OneDrive (or BaiduYun). Or you can train your own word vectors(dimension = 100) using the word2vec method in gensim library.