White Paper Addendum - Plato, the Cave and Recorded Future
By Jason Hines on March 18, 2010
To understand the philosophy behind Recorded Future, it is helpful to consider the famous “cave allegory” by Plato:
Plato imagines a group of people who have lived chained in a cave all of their lives, facing a blank wall. The people watch shadows projected on the wall by things passing in front of a fire behind them, and begin to ascribe forms to these shadows. According to Plato, the shadows are as close as the prisoners get to seeing reality. He then explains how the philosopher is like a prisoner who is freed from the cave and comes to understand that the shadows on the wall are not constitutive of reality at all, as he can perceive the true form of reality rather than the mere shadows seen by the prisoners. (en.wikipedia.org/wiki/Allegory_of_the_Cave)
(image from: www.thatmarcusfamily.org/philosophy/Course_Websites/Phil_Math/Photos/Cave.jpg)
What we read in newspapers, blogs etc. is not unlike the shadows on the wall of the cave – we get reports about events in the real world, and attempt to use that information to get an idea about what is really happening. As good analysts, we naturally consult several sources, and weigh together the information obtained from them – always keeping in mind that some sources are more credible than others, and thus should be given higher weight. We call the evidence we get from different reports “event instances”, and the real world events they report on we refer to as “canonical events”.
A canonical event, in our system, is a representation of a particular happening in the real world. For example, assume we read the following statement in the New York Times:
“Barack Obama said yesterday that Hillary Clinton will be travelling to Haiti next week”
This statement describes two events: a canonical “Quotation” eventand a canonical “Person Travel” event.
The quotation event refers to a canonical entity, “Barack Obama”, and a statement “Hillary Clinton will be travelling to Haiti next week”. It has an associated time, “yesterday”.
The “Person Travel” event includes references to two canonical entities, “Hillary Clinton” and “Haiti”, and has an associated time “next week”.
Note that “yesterday” and “next week” are relative times, and to place them on an absolute time axis we need to know when the entire statement was uttered. Let us assume that the statement was uttered on Wednesday, March 17th. Then we might represent the statement pictorially in the following way:
In our system, this statement will be represented in the following way:
We have three canonical entities: Barack Obama and Hillary Clinton, which are Person entities [blue rectangles], and Haiti, a Location entity [green rectangle].
There are two canonical events [red ovals] – “Quotation by Barack Obama” and “Person Travel of Hillary Clinton to Haiti”.
Furthermore, there are instances of these events [pink ovals], which are tagged by the time or time interval during which they are expected to have occurred or will occur.
The Quotation instance also has a reference to the text of the quote and to the instance of the event referenced in the quote.
Finally, both instances refer to the text fragment representing the original statement, and the fragment refers to its source – the New York Times.
Multiple text documents, retrieved from different sources, can of course be used to gather evidence of the same canonical event, i.e., to provide different instances of the canonical event. Several different canonical events – and instances – will also refer to the same entities. To extend our example, let’s add the statement:
“Hillary Clinton to meet with Ban Ki-Moon in Port au Prince on March 23rd”
The representation of our “world knowledge” will then be updated to:
Is this all we know? Not really! Recorded Future also maintains an ontology, with additional information about canonical entities and their relationships. In this particular example, the following information can be found in our database:
Combining the information derived from analyzed text and the ontology gives us the following picture for this minimal example. In the real Recorded Future database, there are millions of event instances. This should give you an idea about how the richness of Recorded Future data can help you in analyzing events in the real world!
Ontology is the philosophical study of the nature of being, existence or reality in general, as well as the basic categories of being and their relations. Traditionally listed as a part of the major branch of philosophy known as metaphysics, ontology deals with questions concerning what entities exist or can be said to exist, and how such entities can be grouped, related within a hierarchy, and subdivided according to similarities and differences. (http://en.wikipedia.org/wiki/Ontology)