What value does semantic AI bring to Tridion?

The core role and function of a CMS is to help maintain, manage, and create content. Tridion excels at this, enabling both web content and highly structured content to be managed across the entire business, irrespective of the location, language, or department. Further, Tridion’s headless capabilities deliver this content seamlessly across all devices.

 

However, an area where traditional CMS inherently face limitations is in organizing and searching for content. This is where semantic AI comes into the picture, adding a layer of intelligence that delivers user-centric search results and personalized recommendations.

Let’s consider a typical scenario; content is created and stored in the CMS all the time by several people across various departments and locations in different languages. Typically, there is poor naming consistency and inconsistent tagging and content classification which makes it really hard to locate information and reuse it, often leading to content being recreated. This is not a very intelligent approach to managing content.  

 

The first part of the solution is to adopt a taxonomy; you can build one yourself or use a standard one for your industry and adapt it where needed. In both cases you can use so called ‘corpus analysis’. This is a machine learning text-analytics approach to scan a lot of your content and extract relevant concepts and terms.

 

A ‘concept’ holds the meaning, while the ‘terms’ are the (many) words to describe the concept (i.e. synonyms), even in various languages. These concepts and terms can be added to your taxonomy and used to classify content, to suggest content relations and for search purposes.

 

With a taxonomy, you now have a predefined way of tagging (classifying) your content which brings better consistency. Your authors can pick relevant terms from the taxonomy and apply those to content. In Tridion you can manually tag content, but it’s much more convenient to let semantic AI do the work for you automatically. Tests show that this has an accuracy of more than 90%. A mix is of course also possible, where the automated tags are reviewed by a human and only then stored into the system.

 

Once content is classified with terms from the taxonomy, suddenly search becomes much more powerful. The taxonomy knows the terms and their synonyms (through the notion of ‘concepts’), meaning you don’t have to use the exact word that shows up in the content to still be able to find it. It can even match multilingual terms so you can search for instance in English, and get a relevant Spanish result.

 

As a result, when a user enters a query, semantic AI understands the ‘user’s intent’, goes in, and pulls out all the relevant information that is related to that search. It can even be used to share personalized recommendations for the user to explore further. As a result, the user quickly achieves the set goals with semantic search and recommendations. 

 

Semantic AI helps all stakeholders that interact with content managed in Tridion. It’s about anyone interacting with the organization and looking for the content they have been provided access to: customers, partners and employees. Thus, the improved experience is widespread and touches every aspect of the business, driving efficiencies (time and cost-saving) and boosting revenue (higher sales conversion) at the same time. 

 

With the semantic AI layer, Tridion delivers on the most pressing and challenging needs of our customers:

 

Click here to learn more about our Semantic AI capabilities.