Publications
Conceptual thinking
Developing advanced software solutions comes with conceptual thinking and rethinking existing concepts. These concepts have been presented at conferences and papers are available for download on this page.
Change Tracking in the XML-Based LiXuid Production Workflow Using Fonto’s Document History
Journal Article Tag Suite Conference (JATS-Con) Proceedings 2020/2021, Charles W. O’Connor and Bert Willems
April 2021
When Aries was looking for an XML editing platform to build on for the LiXuid XML workflow, a top priority was change tracking that was robust and usable to satisfy the needs of journal production staff. The solutions available at the time, based on using processing instructions, either did not capture the full range of possible changes or were unreliable in their behavior. Aries also wanted to avoid having to superset JATS to add change tracking elements. Fortunately, Fonto, one of the candidate XML editors, was working on a solution based on DeltaXML called Document History. What sets Document History apart is its ability to display changes made across different revisions of an XML document. This principle is called a changelog and it is different from A/B comparisons typically employed in change tracking solutions. A changelog is a comparison between multiple, subsequent revisions of an XML document merged back into one annotated XML document which is then visualized. This changelog contains the necessary information to attribute changes to specific users and moments in time, something that is typically only offered by active change tracking systems. Another property of this changelog is its ability to show overlapping and conflicting changes. Both textual and XML changes are displayed in Document History as a redlined version of the document. Through the UI, users have the ability to determine what range of revisions they want to look at, and the application provides the ability to navigate quickly to the associated XML editor as well as to mark change as seen.
A note on Editor performance – A story on how the performance of Fonto came to be what it is, and how we will further improve it
XML Prague 2020, Stef Busking and Martin Middel
February 2020
This paper will discuss a number of key performance optimizations made during the development of Fonto, a web-based WYSIWYM XML editor. It describes how the configuration layer of Fonto works and what we did to make it faster. It will also describe how the indexing layer of Fonto works and how we improve it in the future.
How to configure an editor – An overview of how we built Fonto
XML Prague 2019, Martin Middel
February 2019
In 2012 a web agency took on the challenge of building an XML editor. This paper gives an overview of a number of concepts that proved to be useful, and some concepts that did not.
Modern amendment drafting – The road to XML
Markup UK 2018, Bert Willems
June 2018
This paper reports on an experiment to build a system which aids in the drafting of amendment documents. The system provides a mechanism to help validate the correctness of amendments. Furthermore, the system is able to semi-automatically sort the amendments in voting order and simulate the effects of amendments on the law. The proposed implementation is based on XML technology, an XML editor and machine learning.
Assisted Structured Authoring using Conditional Random Fields
XML Prague 2018, Bert Willems
February 2018
Authoring structured content with rich semantic markup is repetitive, time consuming and error-prone. Many Subject MaĴer Experts (SMEs) struggle with the task of applying the correct markup. This paper proposes a mechanism to partially automate this using Conditional Random Fields (CRF), a machine learning algorithm. It also proposes an architecture on how to continuously improve the CRF model in production using a feedback loop.
Bridging the gap between knowledge modeling and technical documentation – Engage subject-matter experts to contribute to knowledge management and help them write accurate & correct documentation
XML London 2017, Bert Willems
June 2017
This paper describes an architecture which allows subject matter experts and the systems to co-create both structured content and knowledge models. The proposed architecture creates a knowledge model from structured content which, in turn, is queried to validate and improve the accuracy and correctness of structured content leveraging the expertise of the subject-matter expert. The proposed architecture effectively describes a feedback loop.
Soft validation in an editor environment – Schematron for non-technical users
XML Prague 2017, Martin Middel
February 2017
To allow the use of Schematron in a quickly changing environment like an editor, understandability is crucial. An author may not be an XML expert, so they must be guided in resolving the messages generated by Schematron. A good understandability rests on two pillars: performance and user interface. An author needs constant feedback on the current state of the soft validation. The author must know which places in the document need attention and how to resolve them. The report must then update as soon as possible to enable the author to see the result of a modification they just made. To ensure good performance, a number of technical problems have been solved, this includes a novel dependency tracking system.