3.4. Absence: How can a computer reach this understanding if not all the data is present?

This is perhaps the most interesting quality of data ontologies. Arts, humanities and social science research is premised on an analysis of the evidence at hand, whilst knowing that our sources are only giving us a partial picture because they are often incomplete as testimonies. Historical inquiry, for example, is determined by the completeness of the historical record. 

An ontology enables us to predict or assume information which is not present in our sources because it presents a model of the real world rather than a model that is limited to what is contained within the sources themselves.

For example, let us imagine that our ontology states that all people have a gender, but a trial in the Old Bailey Proceedings Online refers to a defendant who has a name that is not gender-specific: Chris. How can a computer guess the defendant’s gender? If all the other defendants in the Old Bailey have been given a gender and they have also been related to types of crime through the ontology, we can use the ontology to guess Chris’s gender based on gender patterns observed in the relationships between the other defendants and their crimes.

Here is another example, this time from Beyond the Multiplex. Our ontology presents two options for viewing films when a person lives in a town: at home and at the cinema. Sarah says that she lives in Hull and only watches films at home using digital streaming services, whereas David says that he lives in Hull and watches films at the local cinema. A computer can therefore infer that Sarah could watch films at the local cinema in Hull but chooses not to, for whatever reason. The important point here is that Sarah never mentions the local cinema during her interview, but using the ontology a computer can infer its existence as an option because David mentions it in his own interview.

previous page | next page

URL for this page
https://www.dhi.ac.uk/books/ontology-guide/not-all-the-data-is-present