Datasets & Replication Packages
Datasets & Replication Packages
Embedded Files
This page provides access to datasets, replication files and supporting research materials associated with my published and ongoing research.
Where possible, I make the underlying data and code used in my research publicly available to support transparency, reproducibility and further research. Individual entries below provide details of the associated paper, data sources and available replication materials.
Supervised Machine Learning Patent Classification Materials
Supervised Machine Learning Patent Classification Materials
This repository contains the Python-based machine learning pipeline used to classify historical British patent titles. Employing Non-Negative Matrix Factorization (NMF) for topic modelling and text analytics, this code processes digitised patent titles to classify inventions into a distinct set of time-invariant technology categories.
System Requirements
The code is written in Python. Users will need a standard Python environment configured with libraries commonly used for text analytics and data manipulation (e.g., scikit-learn, pandas, numpy, and nltk or spacy). Exact version dependencies are listed within the package documentation.
Citation
If you use this code or the taxonomy in your own research, please cite the corresponding publication:
Billington, S. D. and Hanna, A. J. (2021) '"That's classified!" Inventing a new patent taxonomy', Industrial and Corporate Change, 30(3). Available at: https://doi.org/10.1093/icc/dtaa049
Unsuccessful Patent Applications
Unsuccessful Patent Applications
This repository contains the primary dataset and supplementary appendices used in the analysis of unsuccessful patent application outcomes. It introduces a newly digitised dataset of historical English patent applications from 1783 to 1834, enabling empirical analysis of patenting success rates during the First Industrial Revolution.
Citation
If you use this dataset in your own research, please cite the corresponding publication:
Billington, S. D., & Lane, J. (2026). ‘Success and failure in England’s patent system: New evidence from patent applications, 1783-1834’. The Economic History Review.
Woodcroft Subject-matter Index
Woodcroft Subject-matter Index
This repository contains the Woodcroft subject-matter patent classes and subclasses assigned to English patents, covering 1617-1852. It also contains the Woodcroft subject-matter Index (WSI) which counts the number of classes each patent is assigned to.
Citation
If you use this dataset or the WSI index in your own research, please cite the corresponding publication:
Billington, S. D., & Lane, J. (2026). ‘Procuring Promising Provisions: the British Patent System and the Navy Proviso, 1794-1831’. European Review of Economic History, 30(1). https://doi.org/10.1093/ereh/heaf011.
Page updated
Google Sites
Report abuse