Structured content and semantic AI adoption in practice

Structured content and semantic AI adoption in practice Continuing with the Tridion’s webinar recently, this is the third in a series of blog posts of the event ‘Structured Content Authoring and Semantic AI’. In the previous two blogs, we discussed the need for structured content and semantic AI and adding structure to traditional content. 

In this third blog, Jan Benedictus from RWS’s Fonto division discussed how traditional content could be converted into structured content using semantics. And then deploying artificial intelligence (AI) to deliver the par excellence user experience that businesses aim for but are unable to. 

To recap, Jan brings about 30 years of experience in the online and digital publishing domain. He founded Fonto in 2014 to make structured content authoring available to everyone. RWS acquired Fonto in March 2022 to bring its expertise and user-friendly structured content creation, editing, and review tools to RWS and Tridion customers. 

The documents to data journey 

Jan discussed the three stages of the ‘documents to data’ journey. He emphasized that the first step is all about collaboration to achieve efficient authoring, which should include review and approval of the component levels rather than at the document level. 

The second step is about re-using the data from components to produce content such as documents, e.g., a recurring report related to compliance or marketing. The output is still in traditional form, e.g., a document. 

The third and final step is about pulling data from different components and publishing it in not just traditional form but a range of forms, including digital formats. This is the stage where data and content are integrated and published in different forms for different purposes. 

Three stages in the documents to data journey 

To reuse the data, Jan mentioned that it is critical to store data in a central data source and then, with the correct schema, it can be reused to produce multiple contents. And it is also the right step towards digital publishing. This helps capture semantics from the database and deliver tailored content to users in digital formats. To demonstrate this, he contrasted the traditional way of reusing data vs. using a central repository. 

Traditional vs. modern way of re-using data 

Traditional vs. Modern Jan briefly touched upon Gartner’s hype cycle to indicate and compare the maturity level of structured content and artificial intelligence. Both seem to be at similar maturity levels, with structured content at a slightly higher maturity level when specifically compared to semantic AI. 

Implementing semantics and the role of authors 

Jan briefly introduced Fonto’s solution in the space of semantic AI, which includes four tools: Fonto Editor, Fonto Content Quality, Fonto Document History, and Fonto Review. These align with Fonto’s goal of ‘making structured content feasible for everyone’. 

Fonto product family

To demonstrate the role of such tools, Jan brought our attention to a tool, ‘Clippy’ Office Assistant. It assists the editor in formatting the document. Fonto’s tools help the editors to connect it to the database or artificial intelligence, thus providing semantic assistance. 

Fonto’s Semantic Assistance Tool 

Above is an example of an author in the pharmaceutical sector. Fonto’s assistant recognizes the medicine Acetaminophen and displays the relevant information in a callout box. The box prompts the author/editor to tag it as a pharmaceutical product. This makes it extremely easy for the authors to perform semantic tagging as they can simply click on the tag button for the content to be semantically tagged and linked to that reference. 

Fonto offers many other functions, such as warning authors if the text is vague, identifying generic terms and suggesting specific terms (e.g., if the author types ‘salt’ then prompts to specify the type of salt), etc. 

Managing the new evolved role of authors 

Jan mentioned that the new role/expectation often leads to authors. The reason being it’s a whole different way and method of building content or document. Organizations first need to build a library of components from which authors will assemble a working document and then make edits to it. This then needs to be reviewed, approved and published in different forms. 

New roles of authors 

This process leads to a shift in authors’ roles compared to how they used to do it traditionally. As illustrated above, it involves building a model document/format and then creating a working document using data, structure and semantics. This is more of a centralized role as this works as a framework for the authors or subject matter experts to work with. 

Many times authors do not work with the standard format of the documents. In this new role, they need to be cognizant of and work with the format, structure, and semantics. Fonto has automated the task of generating documents in the desired format and design. Hence authors do not need to work on that aspect unless there is a need for a human element, e.g., in case of marketing or education content. 

All said and done, this way of working is quite different from the traditional way; however, now that the technology is there, Jan is optimistic that authors will adopt and welcome the change gradually as they realize the benefit. 

On a closing note, Jan reiterated the difference between perception and reality about structured content. Forms, templates, ToCs, or standardized sections are often understood as structured content, whereas in reality, it is about schema and structured XML. J

an mentioned that many tools exist for ‘Document Automation,’ which is good. However, the real value lies in technology that equips organizations to publish data in various digital forms. Jan reminded us about the underlying change and Forrestor’s prediction that a mainstream paradigm shift in document production and consumption will happen by 2027. 

For more details, listen to the webinar.