As described elsewhere, the core of the Action Intelligence (AI) Framework is the classification of data (extracted during the AI discovery stage) so that it can be more easily visualized (as part of the AI Analysis stage). Conceptually, at least, classification is applied in two steps:
- Arranging data consistently according to some knowledge representation standard (we’ve created AAAPT to help with this, which stands for Actors, Actions and Artifacts in Place and Time) and
- Applying an interpretative lens to the Actors, Actions and Artifacts according to the needs at hand (we’ve created and used 3by5 sociocultural and 3byX sector lenses to help with this).
At its simplest, AAAPT is designed to classify three types of entities (Actors, Actions and Artifacts) and specify the relationships that link them. A key principle of Action Intelligence is locating these entities in time and space. While each entity type can be further broken down (e.g. an Actor can be further classified as an organisation or a person), this is unnecessary.
An example is worth a thousand blog entries. Let’s look at how AAAPT would be used with the following extract from a news story on the World Health Organization website about the response to the Ebola outbreak in West Africa in 2014. We’ve highlighted the key Actors, Actions, Artifacts, Places and Dates:
“Kailahun district borders the area of Guinea where the first cases of Ebola were confirmed at the end of March 2014…… WHO started working closely with the Ministry of Health of Sierra Leone to prepare for possible imported cases of Ebola from Guinea……WHO immediately established an office in the city…. In June, Médecins Sans Frontières (MSF) established an Ebola treatment centre and WHO helped deploy a mobile laboratory from Public Health Canada to test blood and swab samples inside the MSF treatment centre.”
(Source – http://www.who.int/features/2014/kailahun-beats-ebola/en/)
With this short extract, we see several Actors (WHO, Ministry of Health, MSF and Public Health Canada), Actions (Treatment Centre), Artifacts (Mobile Laboratory), Places (Kailanun district, Guinea, Sierra Leone), and Times (March and June 2014).
While we haven’t yet identified the relationships linking these entities, value can still be derived from this simple classification process.
“How?” I hear you ask.
Imagine repeating this on hundreds of Ebola news stories over the 2014 – 2015 period and storing just the above information in a database (i.e. who, what, where and when). It’s now time to assess and evaluate the effectiveness of the Ebola response.
Most evaluations would ‘start’ with a desk review, with weeks spent reading reports, documents, media releases, etc. and combining hand-extracted notes with other numerical data one might have (costs, numbers treated, etc.). Only then would one then be in a position to begin analysis.
But with AAAPT classification (in its simplest form) and a database (even spreadsheets will do as we’ve demonstrated in low resource settings), one can move much more quickly to analysis. Assuming one has collected and applied AAAPT to news articles as they emerged, then we could immediately move to the initial analysis. For example, we could visualise the arrival sequence of Ebola response actors on a timeline. We could quickly visualise the location of treatment centres on a map and overlay that with the outbreak’s spread. We could compare the response timelines of different Actors. We could …… .
The potential questions are endless, but applying a classification standard (like AAAPT) to the realm of human action is at the core of the AI Framework and is the foundation upon which everything else depends.
In another post, we’ll look at the relationships that link Actors, Actions and Artifacts.
Leave a comment