Implementation of the BISON methodology to predict content success on the web3 platform Mirror (see related publication: https://dl.acm.org/doi/full/10.1145/3748699.3749796).
This repository contains a full case study of Writing NFTs published on Mirror, using NLP techniques, blockchain feature extraction, machine learning and explainability techniques. The project explores how textual and blockchain-related features can be used to predict content success on decentralized social media ecosystems.
The original dataset WNFTM is provided in a reduced form of 100 Writing NFTs (mirror_publications_100.csv). For access to the complete dataset, contact: framay909@gmail.com.
For a full description of the dataset features, refer to: https://github.com/LauraPollacci/BISON.