Vasari Social Network

Index

Introduction

This website contains all the documentation related to the extraction of a Social Network from Giorgio Vasari's The Lives of The Artists (1568). Aim of this work is to use Information Extraction techniques, such as Named Entity Recognition (NER) and Coreference Resolution (CR) to identify references to individuals inside the book and compute connectedness between persons by measuring their statistical association using paragraph co-occurences.

In order to identify the references to artists and craftsmen, the Index of Names is used, containing for each artist the list of pages in which he/she appears. By using this bibliographic information in tandem with NER and CR, it is possible to identify each occurence of the artists in the paragraphs of the pages.

Finally to measure connectedness between artists, a measure of statistical association called Pointwise Mutual Information (PMI) is used. This measure quantifies how much more likely these artists occur together than if they were independent, statistically speaking. In simple terms, PMI tells us how much we can learn about one artist by knowing that another one is referenced in the same paragraph. If the PMI is positive, it indicates a positive association, meaning the artists tend to occur together more often than expected by chance. We used PMI to filter out associations with a negative value and to weight the connection between artists inside the biographies.

This social network was extracted from a 10 volumes edition of Vasari's work, translated in English by Gaston du C. De Vere. In this website we provide a dynamic visualization of how the social network of artists evolves across all 10 volumes.

Knowledge Graph

In order to share the final results in a FAIR data format, an RDF dataset of the Vasari Social Network is made available:

The RDF data is modelled by using standard vocabularies such as RDFS, SKOS and FOAF. A visualization of the conceptual model is presented in the image. All edges are represented as blank nodes, while each node has several attributes related to its network centrality.

Data Model of the Knowledge Graph

Dynamic Network Visualization

The following links provide graph visualizations produced from each volume of Gaston du C. De Vere's translation of The Lives of The Artists. The visualizations are produced with Pyvis.

Resources

This section lists datasets and source code related to this project:

Contacts

If you are interested in the project and want to get in contact, you can reach one of the contributors below: