When will the pandemic be over? Computer science may have the answer

In early 2022, practically two a long time following Covid was declared a pandemic by the Environment Wellbeing Group, gurus are mulling a massive query: when is a pandemic “over”?

So, what’s the respond to? What requirements should really be employed to decide the “end” of Covid’s pandemic stage? These are deceptively very simple inquiries and there are no easy solutions.

I am a laptop scientist who investigates the development of ontologies. In computing, ontologies are a suggests to formally structure knowledge of a subject domain, with its entities, relations, and constraints, so that a computer system can process it in various purposes and help humans to be far more specific.

Ontologies can find expertise that’s been overlooked till now: in 1 instance, an ontology recognized two additional useful domains in phosphatases (a group of enzymes) and a novel domain architecture of a portion of the enzyme. Ontologies also underlie Google’s Know-how Graph that is guiding all those expertise panels on the ideal-hand side of a look for end result.

Making use of ontologies to the issues I posed at the start is beneficial. This approach will help to clarify why it is complicated to specify a lower-off position at which a pandemic can be declared “over”. The approach requires collecting definitions and characterizations from domain industry experts, like epidemiologists and infectious condition experts, consulting applicable exploration and other ontologies, and investigating the nature of what entity “X” is.

“X”, in this article, would be the pandemic by itself – not a mere shorthand definition, but seeking into the properties of that entity. Such a exact characterization of the “X” will also reveal when an entity is “not an X”. For instance, if X = residence, a house of residences is that they all will have to have a roof if some item doesn’t have a roof, it undoubtedly is not a home.

With those people attributes in hand, a precise, official specification can be formulated, aided by further methods and equipment. From that, the what or when of “X” – the pandemic is over or it is not – would logically observe. If it does not, at the very least it will be feasible to describe why factors are not that simple.

This sort of precision complements wellness experts’ initiatives, aiding human beings to be extra exact and connect much more specifically. It forces us to make implicit assumptions specific and clarifies where by disagreements may be.

Definitions and diagrams

I performed an ontological examination of “pandemic”. Initially, I needed to obtain definitions of a pandemic.

Informally, an epidemic is an prevalence throughout which there are numerous situations of an infectious condition in organisms, for a limited duration of time, that has an effect on a neighborhood of mentioned organisms dwelling in some location. A pandemic, as a minimum amount, extends the area where by the infections acquire spot.

Future, I drew from present foundational ontologies. This has generic categories like “object”, “process”, and “quality”. I also utilised area ontologies, which comprise entities distinct to a matter area, like infectious disorders. Among other sources, I consulted the Infectious Condition Ontology and the Descriptive Ontology for Linguistic and Cognitive Engineering.

Very first, I aligned “pandemic” to a foundational ontology, using a determination diagram to simplify the method. This helped to function out what variety of matter and generic class “pandemic” is:

(1) Is [pandemic] a little something that is taking place or occurring? Indeed (perdurant, i.e., a little something that unfolds in time, instead than be wholly current).

(2) Are you equipped to be existing or take part in [a pandemic]? Certainly (event).

(3) Is [a pandemic] atomic, i.e., has no subdivisions and has a definite endpoint? No (accomplishment).

The term “accomplishment” may possibly seem to be unusual here. But, in this context, it will make apparent that a pandemic is a temporal entity with a limited lifespan and will evolve – that is, stop to be a pandemic and evolve back again to epidemic, as indicated in this diagram.

Properties

Upcoming, I examined a pandemic’s traits described in the literature. A extensive checklist is described in a paper by US infectious condition experts released in 2009 during the worldwide H1N1 influenza virus outbreak. They collated eight attributes of a pandemic.

I stated them and assessed them from an ontological viewpoint:

1. Extensive geographic extension. This is an imprecise element – be it fuzzy in the mathematical sense or believed by other implies: there isn’t a crisp threshold when “wide” begins or ends.
2. Disorder movement: there’s transmission from spot to position and that can be traced. A of course/no attribute, but it could be manufactured categorical or with ranges of how bit by bit or quick it moves.
3. High assault fees and explosiveness, or: a lot of persons are impacted in a brief time span. Quite a few, small, fast – all suggest imprecision.
4. Negligible inhabitants immunity: immunity is relative. You have it to a diploma to some or all of the variants of the infectious agent, and also for the population. This is an inherently fuzzy characteristic.
5. Novelty: A indeed/no feature, but a person could include “partial”.
6. Infectiousness: it should be infectious (excluding non-infectious things, like obesity), so a very clear of course/no.
7. Contagiousness: this may perhaps be from individual to particular person or by means of some other medium. This home incorporates human-to-human, human-animal middleman (e.g., fleas, rats), and human-setting (notably: drinking water, as with cholera), and their attendant factors.
8. Severity: Traditionally, the term “pandemic” has been utilized additional usually for extreme diseases or people with substantial fatality fees (e.g., HIV/AIDS) than for milder types. This has some subjectivity, and thus may perhaps be fuzzy.

Properties with imprecise boundaries annoy epidemiologists since they may possibly lead to unique results of their prediction models. But from my ontologist’s viewpoint, we’re acquiring somewhere with these properties. From the computational side, automated reasoning with fuzzy characteristics is attainable.

COVID, at minimum early in 2020, conveniently ticked all 8 bins. A suitably automated reasoner would have categorized that scenario as a pandemic. But now, in early 2022? Severity (place 8) has mostly reduced and immunity (level 4) has risen. Point 5 – are there worse variants of worry to occur – is the million-dollar dilemma. Far more ontological evaluation is essential.

Highlighting the issues

Ontologically talking, then, a pandemic is an function (“accomplishment”) that unfolds in time. To be categorised as a pandemic, there are a number of characteristics that are not all crisp and for which the imprecise boundaries haven’t all been established. Conversely, it implies that classifying the occasion as “not a pandemic” is just as imprecise.

This isn’t a entire response as to what a pandemic is ontologically, but it does lose mild on the troubles of calling it “over” – and illustrates very well that there will be disagreement about it.

This post by Maria Keet, Associate professor in Computer system Science, University of Cape City is republished from The Conversation underneath a Imaginative Commons license. Go through the primary short article.