Modelling Semantic Diversity in Knowledge Economy
Repository to accompany the article: "Misalignment between Skills Discovered, Disseminated, and Deployed in the Knowledge Economy", Bhargav Srinivasa Desikan and James Evans.
This repository contains the cleaning code, model creation and hyperparameter details, and some statistics on the make-up of the data. The notebooks folder contains all the jupyter notebooks used for analysis. The images folder contains all the figures and plots from the paper as well as more figures and plots not included in the paper.
Organisation
The notebooks are organised as belonging to either:
- data cleaning, metadata, and statistics
- embedding creation and vector aggregation
- data analysis (scaling and statistics)
These notebooks cover all the analysis done presented in the paper, while also including more information on the data and analysis not included in the paper.
Data
The data folder has some samples of the data. The embedding models and vectors for research, teaching and industry can be shared by contacting the repo owner.