20 July 2018
Freddie McMahon


Freddie McMahon - DF2020 Ltd

46Posts 256,379Views 2Comments
Finance 2.0

Finance 2.0

A community for discussing the application of Web 2.0 technologies to financial services.


05 January 2018  |  5893 views  |  0

Just as a reminder, Complex Knowledge is any combination of regulatory, statutory, legal, tax, tariff, policy and procedure matter, which is primarily found within documents.




Preparing the Complex Knowledge from Monologue to Dialogue

In context to the earlier defined scope, the knowledge was extracted and synthesised from the two government documents representing 215 pages. This was relatively straight forward as it was from the ‘original source’. The activity mainly involved removing noise (not needed) or repetition by generally keeping the content that contained the best narrative clarity.

The next stage was to create a simplistic algorithmic structure. The very nature of this type of Complex Knowledge suited itself to forming into a decision-tree consisting of Choices, Pathways and Outcomes. Remember, this is knowledge driven and not data driven. There is no database in these documents!


Identifying Decision Risks

The next consideration relates to identifying the risks in context to each Choice.

This might sound an obvious consideration.

However, there are two key points about the status quo of documented Complex Knowledge:

  • Though these documents can be ‘scrupulously’ reviewed and approved, they are rarely subject to usability tests – this is a major failing of documented Complex Knowledge as these two disciplines are quite different.  


  • Even worse, there is no audit trail of the user decision journey through the documented Complex Knowledge, as applied in practice. The absence of a transparent audit trail of the user decision journeys has enabled deep systemic risks to thrive and evolve below the radar of most governance, risks and controls.

Each Choice within the Complex Knowledge algorithm inherently contains a risk, which may result with the wrong Option being selected when applied in practice. It is like driving a car and coming to a T-Junction (2 Options) or a Cross-Roads (3 Options) – selecting the inappropriate Option means following the wrong Pathway leading to a different Outcome. Oddly enough, having too many Options of say six or more increases the risks, which is typically not well understood.

Where the Choice narrative could be misinterpreted then techniques can be deployed to reduce the risk – one of these techniques is covered next.  

Several Choice-based risks emerged within the Use Case for the Loft Extension. Here is one example:


“Is the extension to be higher that the highest part of the roof?”.


This Choice narrative could be misinterpreted.

For example, some people may misinterpret by believing the chimney is the highest part of the roof as opposed to the apex of the roof. 

This risk can be reduced by providing a picture in-line with the narrative that illustrates the roof and its apex along with the chimney. Of course, we all understand that a “picture can be worth a thousand words” to convey a meaning or essence more effectively than a description does.

There are other techniques that can be used to reduce the risks at the granularity of Choice. In this Use Case, pictures were used for selective Choices to reduce decision risks.

None of these pictures were in the 215 pages nor were they found in the other content provided representing processes, procedures and web content.


The Importance of Nano Risks

The term nano risks is used to represent the inherent elements within the abundance of Choices and their Options found within Complex Knowledge documentation. There seems to have been no empirical study that has identified the volume, variety and velocity of nano risks within Complex Knowledge and their impact upon Outcomes within the Public and Private sectors.

The exposure of these nano risks are systemic, which means they have been able to bypass the rigours of signatory approvals and audit sign-offs for Complex Knowledge documentation.

At this stage it is worth reflecting upon nano risks in context to the earlier statement mentioned by Dame Judith Hackitt:

“it has become clear that the whole system of regulation, covering what is written down and the way in which it is enacted in practice, is not fit for purpose ...”

Sadly, Complex Knowledge nano risks has proven politically sensitive in Financial Services. Off the record discussions with banking executives, external auditors and trade bodies has led to a reluctance to go further unless the industry moves as one with the backing of the regulators.  Therefore, a utility could be a saving grace with the justification of not wanting each firm to ‘reinvent the wheel’.


Creating the Knowledge Map and Chatbot

To be continued .. part 4



TagsArtificial IntelligenceRisk & regulation

Comments: (0)

Comment on this story (membership required)

Latest posts from Freddie

Conversation-as-a-Service insights for changing everything

7 hours ago  |  1255 views  |  0 comments | recomends Recommends 0 TagsArtificial IntelligenceStart upsGroupFinance 2.0

The Normalisation of Procedural Subjectivity

06 July 2018  |  4688 views  |  0 comments | recomends Recommends 0 TagsArtificial IntelligenceRisk & regulationGroupFinextra site news

Freddie's profile

job title Chatbot Thought Leader
location London
member since 2017
Summary profile See full profile »
Pioneering RegTech Chatbots http://www.chatbotauthor.ai/ http://www.df2020.com/ http://www.chatbotinsider.ai/

Freddie's expertise

Member since 2017
22 posts2 comments
What Freddie reads

Who's commenting on Freddie's posts

Dharmesh Mistry
Ketharaman Swaminathan
Balwant Surti
Emmanuelle Johaadien