README.md

Matching Accounts on Blockchain with Pseudo Fine-tuning of Language Models

Code and dataset for the submission "Matching Accounts on Blockchain with Pseudo Fine-tuning of Language Models".

Getting Start

Requirements:

• Python >= 3.6.1
• NumPy >= 1.18.1
• TensorFlow >= 1.4.0

1. Preprocess dataset

Step 1. Download dataset from Google drive:

• All_in_one
• Typhoon dataset (new)

Step 2. Unzip dataset under the directory of "PFT/Data"

tar -xvf PFT_Data.tar.gz

2. Pseudo-supervised learning

Please refer to ./Model/run_pplm.sh