A key piece of information which needs to be considered (and has been absent so far) is the relationships between entities in the world.
For instance, in the world of finance, when we want to estimate the value of a manufacturing company, just focusing on the
historical price information (returns, earnings, revenue etc) won't provide the full picture.
If we had an AI system which knew all the products of the company, the components of each product and where they were sourced
from as well as the market share of the company and the demand for its many products we would be in a better position to estimate
the value of this company.
We could go even further and consider entities like public opinion about each product, the impact of various legislatives on them
and the various ways significant events like climate change influences the company as a whole. The list of entities which could
be considered goes on.
It would also be great if our AI system knew the same relations for all the competitors of this company. And it doesn't have
to be limited to a specific sector or industry, it could encompass all the economy.
Notice that it's not just the data we want to have access to, the true value here is in the relations that connects these information.
Hence, what we're talking about here is a knowledge graph where entities are linked to
one another through relations. Each link between two entities in this graph can have a type (like investor, product, impact etc) and a value
representing it's strength.
Automated Reasoning Using Relations
Our knowledge graph contains all the entities and relations of interest. In a way, for each entity it provides its context, defines the world it lives in
and sees the dynamics which make it change and evolve. In other word, it contains the meaning or semantics of the entity.
We humans can reason about things because we have a good picture of the world in our minds and we know how things are related to one another. We derive the meaning of things from this knowledge and through
that we can reason about things.
The same applies to machines.
Including relations in analytics by using knowledge graphs enables us to encode semantics in a machine readable way and this makes Automated Reasoning possible.
Given this ability we can invent and quantify meaningful concepts that everyone is familiar with but no one has ever built before. For example we can, for the first time, quantify
the exposure of each company to a given event (say, elections, pandemics, natural
events). As another example, the automated reasoning engine could calculate the likelihood of a merger success between companies and even identify the
most likely post-merger competitor. We can simulate the impact of a supply chain disruption and redefine age old financial concepts like risk and correlation to improve them.
This type of automated reasoning is precisely what we do in Atlas 4.0, not only for financial assets, but for entities in other sectors as well (sectors such as media and security).
This is because the reasoning behind calculating these concepts for financial assets can be generalized and used for entities like political movements, people of power, countries ...
metrics like political influence score, social impact score and more can emerge.
To find out more about our solutions contact us or ask for a demo.