Global Standards based Accounting and Reporting Oracle Machine Prototype

I keep coming back to the notion of an oracle or oracle machine.  Prior posts are here and here. Now, I have a prototype of what something like an accounting and reporting oracle machine might look like.  First, let me summarize again what an oracle machine is.

An oracle is a person or agency, like a software application, considered to provide wise, insightful, useful information or counsel or perhaps useful simulations or predictions.  For example, a Certified Public Accountant (CPA) provides accounting and business related advice on specific topics; a CPA is an example of an oracle: a trusted business adviser.

An oracle machine can be thought of as a Turing machine connected to an oracle of machine-readable information and rules. The oracle, in this context, is a software application capable of solving some computational problem (logical, mathematical), which for example may be a construction problem, a decision problem, or a function problem. Think rules-based expert system for very precise and accurate information and counsel or maybe a ChatGPT "copilot" type thingy that is more probability based, but helpful.

If you don't understand rules-based expert systems or intelligent agents; it is really time to learn. There is ZERO PROBABILITY that artificial intelligence will have no impact on accounting, reporting, auditing, and analysis. Exactly WHAT impact is still to be determined, but trust me; the impact will not be "nothing".

Here is a diagram that explains what an intelligent software agent might look like and do: (to understand more details, read the information provided above)


Knowledge "is a form of awareness or familiarity. It is often understood as awareness of facts or as practical skills, and may also mean familiarity with objects or situations." (from Wikipedia)  You can think of knowledge as your skills which came from education and/or experience. This knowledge that you accumulate allows us humans to draw conclusions, have insights, come up with ideas, and generate wisdom from information by combining the information, skills, and experience.

A knowledge graph is a technology that seeks to turn data into machine-interpretable knowledge. A knowledge assembly is some set of knowledge graphs. A set of knowledge assemblies forms a knowledge graph system. Knowledge graph systems can be created for financial reporting.

Knowledge assemblies based on global standards can be shared more easily publicly than proprietary knowledge assemblies because the logical models and technical syntax used to create the knowledge assemblies tend to be easier to share which provides leverage. Also, standards make markets. Think how the standard ISO shipping container changed the global marketplace.

Curated knowledge graphs tend to be higher in quality than non-curated knowledge.  Many stakeholders have an interest in global standard machine readable knowledge graphs that are curated and the origin of the information is well understood. The management of knowledge provenance can be very important sometimes; things like blockchain, NFTs, smart contracts and such can be used to manage that using things like digital distributed ledgers.

Paraphrasing from an SAP article that explains knowledge graphs and the machine readable knowledge networks they are used to create well, knowledge graphs allow us to map how we humans understand the world, our skills and experience, our accumulated knowledge, how we humans go through life and how we contextualize information in our minds; and then share that knowledge with others for fun and profit. This ability allows us humans to draw conclusions that produce things like fascinating “aha” moments.  Imagine someone paying you to use your knowledge that you have put into machine-readable form.

Why would you do this?  Why is this useful?

Ultimately, this is also the goal of business software: to connect knowledge and find solutions to problems, ideally in an automated way. Then, this business software can be used to augment the skills of people such as professional accountants.  Such software applications can, using that machine-readable knowledge, "supercharge" an accountant's ability such as identifying exceptions and detecting anomalies and performing rote, mechanical, repetitive work.  It is the knowledge within the machine readable knowledge graphs that make that business software perform what looks like magic.  But it is not really magic; it is the effective use of the machine-readable knowledge contained within a high-quality knowledge graph.

An "oracle machine" is just a fancy name for a knowledge graph with some software attached to it that performs useful work. To make this work, the reasoner needs to be "educated" with curated pieces of explicit knowledge, good practices, rules of thumb, and other such helpful clues to help the reasoner along.  Results need to be trustworthy, interpretable, explainable; the line of reasoning needs to be clear and outlined step-by-step.  The origin of the knowledge used must be documented and auditable.

Here is an example of what can be called a global standards based (i.e. it is all global standard XBRL technical syntax; forthcoming OMG Standard Business Report Model logic) framework for representing knowledge graphs or knowledge assemblies or however you want to refer to them.  This is information that I know, knowledge:

  • Standards (prototype of human readable version of financial reporting scheme, work in progress)
  • Topics (used to organize disclosures)
  • Disclosures (used to report information, can be organized using topics)
  • Structures (used to construct disclosures using blocks of information)
  • Terms (detailed objects used within a disclosure)
  • Fundamental Accounting Concepts (high level terms used to create continuity crosschecks)
  • Reporting Styles (used to provide report creation flexibility; needed because different reporting economic entities can use different patterns of report models)
  • Types-subtypes (a.k.a. wider-narrower, general-special; class-subclass; used to control permitted term usage)
  • Consistency crosscheck rules (used to manage information quality)
  • Derivation rules (used to derive information that was not provided from other information that was provided using known rules)
  • Templates (reporting logical patterns that can be leveraged to create a report)
  • Exemplars (examples of a disclosure provided within some other reporting economic entities report that can be borrowed, similar to a template, to create your report)
  • Financial Reporting Scheme (collection of all of the above stuff used within a reporting system)
  • Report Repository (collection of reports prepared using a reporting scheme by reporting economic entities for some period of time)
An untrained observer will look at the above and think, "This is a toy."  But a knowledgeable, trained observer will see a framework.  What you see above was created using the Seattle Method which is a framework for building a complete XBRL-based digital reporting systems when reporting economic entities are permitted to modify their report model.  The Seattle Method was created by reverse engineering the XBRL-based reports submitted by public companies to the U.S. Securities and Exchange Commission using US GAAP and IFRS (here are some examples) and XBRL-based digital financial reports submitted to the ESMA using IFRS (here are some examples). (This document walks you through the reverse engineering process.)

Here you will find a conformance suite, test cases, business use cases, and a showcase of reports that can be handled by this framework.

Here you can find older, now pretty stale working proof of concepts and prototypes created for US GAAP and IFRS and several other financial reporting schemes: (these are now outdated but are used to show that this framework works for US GAAP, IFRS, and other such robust financial reporting schemes)

Finally, here is software that leverages this framework: (in order that the software was created)

The best summary of the results of my 20 years of sensemaking is summarized in this blog post: Puzzle Pieces of Digital Financial Reporting.  

This is less about "XBRL", more about leveraging the characteristics of XBRL to enable digital financial reporting.  All of these ideas can be leveraged by general business reporting also.  If you look deep and hard; what you will recognize is that what has really happened is that a new, modern approach to the electronic spreadsheet has been created.

An accounting and reporting oracle machine is effectively a decentralized knowledge graph.  It is a corpus of information, a source of truth in machine readable form.

Additional Information:

Comments

Post a Comment

Popular posts from this blog

Getting Started with Auditchain Luca (now called Luca Suite)

Image
(Note that a newer, production version as opposed to only the development version , of Luca Suite is available) The Innovator's Dilemma points out that disruptive innovations often look like toys that are doing everything wrong. Auditchain Luca , which is part of the Auditchain Suite , is a unique approach to creating a knowledge graph .  To experience using Auditchain Luca, have a look at the Getting Started with Auditchain Luca instructions. Be sure to grab the Excel and JSON import files used for this getting started exercise.   This getting started exercise uses one of several different approaches to getting information into a tool for constructing and viewing professional quality knowledge graphs .  Additional getting started guides will be made available over time. Auditchain Luca might not make sense to you if you don't understand the paradigm shift that is occurring and if you have not shifted your mental map. For the next step in your journey, consider t...
Read more

Relational Knowledge Graph System (RKGS)

Image
Relational Knowledge Graph System (RKGS) is an interesting idea that combines the relational paradigm and the knowledge graph paradigm.  This documentation, Why RKGS , explains the idea. This gets even more interesting with the notion of "relational AI".  RelationalAI is described as follows by the company of the same name, Relational.ai : “RelationalAI is a cloud-based relational knowledge graph management system, with state of the art probabilistic processing and declarative reasoning at scale to make developing Data Applications a superpower for your business.” This notion of knowledge graphs represented in relational databases fits into my belief that knowledge graphs have two aspects: the expression SYNTAX and the logic of what is expressed or SEMANTICS. While the LOGIC of a knowledge graph is always the same regardless of technical syntax used to express that logic; there are many, many different technical syntaxes that can be used to represent a knowledge graph. There...
Read more

Professional System for Creating Financial Reports Leveraging Knowledge Graphs

Image
The document Professional System for Creating Financial Reports Leveraging Knowledge Graphs explains a new approach, a new paradigm, for creating financial reports.  The document provides a practical, nontechnical description of a professional system for creating provably high-quality financial reports where the report creator is using financial reporting schemes (such as US GAAP and IFRS) where report creator is permitted to modify the report model which leverages global standard knowledge graphs represented using XBRL. This new approach is useful, it is highly novel, and the details were highly complicated to pull together.  This is certainly not an incremental innovation.  It is definitely a disruptive innovation and it could very well be a foundational innovation. You can decide which for yourself. Foundational technologies have two dimensions that need to be considered: novelty and complexity. The more novel the technology is, the more effort needed to ensure users ...
Read more