XBRL-based Digital Financial Reporting

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 fo

I have been thinking about XBRL-based digital financial reporting now for about 20 years.  I collected a lot of really good information and put it into a document, Essence of Accounting . I was trying to explain how financial reporting worked to to someone the other day and I through a lot of "stuff" and I put that stuff on a piece of paper.  To better think through this, I put the information into PowerPoint, and organized it as a very rudimentary infographic.  Here is the result of that brainstorming: ( here is a JPEG of that graphic ) The organization above did not get me where I wanted to go, so I reorganized the information above into this "flow" which I think is better: That is perhaps a bit hard to understand, so here is a narrative that tries to explain what I am trying to get across: Math and Logic (in red): Fundamentally, financial reporting is grounded in the rules of mathematics and logic.  The rules of mathematics and logic precede any rules that standa

It is time for another look at the world's first expert system for creating financial reports. Here are some ways you can take a look should you be interested: Auditchain Luca Demo Creating Financial Reports  (Video) Take Auditchain Luca for a test drive (realize that the software is still under development Showcase of Reports created using Auditchain Luca Test cases, conformance suite, business use cases There is a lot there to see; but you really need to understand what you are looking at. Think about something: Exactly how much does Microsoft Word or Excel understand financial reports or financial accounting?  Zip, zero, nota.  But what if software did, in fact, understand financial reports to a useful degree?  How exactly would you communicate that information to such a software application?  Here is how: Provide a logical model. Here is my logical model, Logical Conceptualization of Financial Report . Provide a framework that leverages the model.  Here is my framework, the S

The contemporary approach to representing a financial reporting scheme or the conceptual framework of a financial reporting scheme has been to publish that information in a book.  For example, AASB 1060 General Purpose Financial Statements – Simplified Disclosures for For-Profit and Not-for-Profit Tier 2 Entities and the related  conceptual framework are represented in that manner. But is there a better way?  Well, this academic paper, An analysis of fundamental concepts in the conceptual framework using ontology technologies , thinks there is.  Other papers think this is a good idea also, see the "Additional Information" section below. And I also think this is a good idea. In fact, I have been representing that sort of information using XBRL for a number of years now.  Here is an example of the key aspects or essence of  AASB 1060 represented using XBRL .  All that information is machine readable and it can be rendered informally in human readable form. Something like an o

The Federal Energy Regulatory Commission (FERC) which switched from a proprietary report information collection system based on Foxpro to XBRL a few years ago is considering further modifications to their information collection system which would include a move to XBRL-CSV. You can see the notice of proposed rule change here on the FERC web site.  Here is additional information provided by XBRL US . XBRL-CSV is part of XBRL International's Open Information Model 1.0.  

DITA or the OASIS Open Darwin Information Typing Architecture  is a standard XML-based architecture for representing documents intended primarily for consumption by humans. The DITA specification defines a set of standard document types for authoring and organizing documents using topic-oriented information and a set of mechanisms for combining, extending, and constraining document types. DITA's architecture is driven by two overarching requirements: Enable interchange and interoperation of content from a wide variety of sources without requiring everyone involved to agree on one single standard document type definition. Enable reuse of content among different publications and/or within the same publication. DITA is not a single document type or application. Rather, DITA it is a framework and set of building blocks by which specific applications can be built while ensuring interoperation and interchange among all conforming DITA content sets and all conforming general-purpose DIT

This post captures thoughts related to XBRL and PDF.  On the one hand, XBRL is a machine readable format; but the information also needs to be readable by humans.  On the other hand; converting XBRL reports and report models can be completely automated.  Not judging one way or another, here are some thought on XBRL and PDF: XBRL to PDF Converter : This tool can be used to convert XBRL to PDF. Now, I have not gotten this to generate a human readable version of a machine readable XBRL report; but perhaps this idea is a good one. ( Here is another XBRL to PDF converter. ) PDFHost : This is a free tool for putting a PDF file somewhere on the internet.  It seems quite easy to use.  You can EDIT a file .  You can make a file READ ONLY .  An issue with this site is that it appears that anyone with the link to a document can edit or even delete the document. Generating PDF from XBRL : This is a PDF file that was generated from XBRL. XSL Formatting Objects (XSL-FO) : This is a W3C standard lang

On a conference call today, Flores Bakker , Enterprise Architect at Ministerie van Financiën in the Netherlands, demonstrated the public tooling he has been working on, including OntoMermaid ("a RDF-based vocabulary to express OWL ontologies into Mermaid diagram language") and htmlvoc ("A RDF-based representation of the HTML-vocabulary to express HTML-documents in RDF, rendering them semantic."). This article, What is Semantic HTML? , provides a graphic that helps you understand semantic HTML by contrasting it to non-semantic HTML: Each creator of an HTML page could use their own semantic HTML, but that wouldn't work out so well for users of those web pages when working across many different web sites.  So, a standard vocabulary of semantic HTML is necessary, and that is being created by the W3C Semantic HTML Vocabulary Community Group . More information about this can be found on the W3C Semantic HTML-Vocabulary Working Group  web page. There is working code o

People are starting to figure out how to apply generative artificial intelligence to accounting and reporting processes.  Here are three examples: Engine B's Copilots :  As explained in this press release , Engine B is bringing "copilots" that leverage generative AI to audit, tax, and accounting markets. Workiva's Generative AI : Watch the video about half way down the page.  It shows how generative AI can be used effectively in the process of creating reports. PWC's Consulting Chatbot : This is a collaboration between OpenAI and PWC to offer consulting on complex issues such as tax, legal, and human resources. Jinsei.ai is a startup that is unifying generative AI, machine-learning, and rules-based AI. There are very likely others creating similar solutions that leverage GPT4 transformers working on specific large language models (LLMs).  Those solutions will very likely prove to provide benefits.  But this will only be the beginning. Imagine combining technologi

Ontology-like things  (a.k.a. knowledge graph systems ) are becoming increasingly important.  Ontology-like things are machine-readable maps that help you understand your data and turn that data into information to make use of that data. It is critically important to get those ontology-like things right because they will drive work in the information age.  As  this LinkedIn post  points out "layering" is important when you construct your " enterprise ontology " and provides this graphic of what layers make up such an enterprise ontology: There are other dynamics also.  Modularity when creating ontologies is important so that pieces can be reused effectively.  Access is also important; some ontologies will be publicly available while other ontologies will be private . Remember; when you think of "ontology", recognize that there are many different approaches to representing "ontology-like things" in machine-readable form: ( larger image ) Quality