We will begin by identifying around 500 individuals who have held power in the 20th Century in one of five leadership domains, namely politics, business, the military, religion, and the professions (including teaching, medicine and the law).
We will create databases containing structured and unstructured information relating to development, education, behaviours, and experiences, derived from a variety of sources.
Unstructured data will be harvested both from online information resources (e.g. Project Gutenberg, Wikipedia, Biography.com, Infoplease.com) and from print and broadcast media. Features relevant to the knowledge graph representation of individual leaders will include their early life histories, professional experience, documented behavioural traits and actions/achievements.
Markers of behaviour at different time points will be based on standardised representations of documented descriptions, but also derived from features of language used in speeches or other official communications, activity on social media and/or in the press, books and articles.
Data will be integrated into a single knowledge graph and analysed using machine and deep learning techniques.
Vidal et al., (2019) define a knowledge graph as ‘the intersection of the formal models able to represent facts of various types and levels of abstraction using a graph-based formalism.’ Using a framework of data ontologies, the KG data format represents entities in terms not only of their properties but also the relationships between them (see Figure 1).
Figure 1, A small section of the Leadership Hubris Knowledge Graph showing examples of the semantic information that can be represented using a simple ontology. Note that Tony Blair, but not Donald Trump, is depicted as possessing the attribute of HS. This is because the requisite knowledge was not available at the time the KG was drawn. After an interval, the KG can be updated as new information becomes available. Alternatively there may be sufficient information present in the full KG to allow the link to be inferred indirectly.
The density and interconnectedness of this data format enable artificial intelligence techniques to be used to characterise the complex relationships among factors responsible for the development of Hubris, and also to deduce new information, generate predictions and test hypotheses about its origins and mechanisms.
It is expected that patterns will emerge which indicate the factors predisposing a leader to (or protecting them from) Hubris Syndrome.
Key terms
View all Close all
Ontologies include abstract concepts or classes – represented as nodes– and predicates representing the relations of these classes –edges in an ontology; the meaning of the predicates is represented using rules.
They are utilised to describe the meaning of the relations, as well as for annotating entities in a uniform way.