Before diving into how AI is/can be related to CMSs, it is important to define boundaries and give some definitions so we can be all on the same page. Depending on who is reading this article, we can have different perceptions of what an AI is or is not and what it can or cannot do. In this article, I’m referring to types of AIs that are related to natural language processing and cognitive computing which is adaptive and contextual. This article doesn’t discuss general intelligence nor AIs that can code its’ components and expand by itself, since it is beyond the scope of this article
It is important to note here that one could consider AI as an empowering tool and not as an adversarial entity cSompeting for our jobs. Therefore, from this point of view we can automate all sorts of tedious tasks and focus on creating quality content, which is one of the main objectives of a CMS (i.e. streamlining content creation). However, a very important thing to note here is that since regulations for using AI are still in their infancy, some might utilize the tools to harm others rather doing good. Therefore, I kindly implore you to use this knowledge for good, not for evil.
AI and Content management systems
When publishing an article, one usually inserts all sorts of images, diagrams and clip arts in order to aid in delivering the content message. For instance, in order to improve the SEO of an article, content authors tend to tag the images which they have with the all possible keywords that identifies that image. Taking writing an article about horses as an example and looking at the following image here:
A photo of an Arabian horse
One might tag the above photo in the following way
Horse, brown, brown horse, horse running in field
Surely, a more seasoned SEO expert might add two or three keywords to the above list. However, if we utilize computer vision and AI on this very image, these are some of the tags we would get:
Horse, Mammal, Vertebrate, Mane, Stallion, Pasture, Mare, Mustang Horse, Grassland, Sorrel, Organism, Livestock, Meadow, Sky, Landscape
Amazing, isn’t it? As a matter of fact, I’ve been tagging images of the articles I’m writing here using this technique, it is easier, faster and mostly accurate. Now imagine a CMS implementation that has this feature at its core, content authors will not need for additional time to analyze photos and/or images in order to tag them appropriately. Instead, they will utilize the APIs through the CMS in order to provide suggestions for the appropriate tagging that might go with a specific image.
Auto/Assisted content authoring
There are many examples of assisted content authoring out there on the internet, even it can be a full-fledged robot such as Bertie. These are kinds of online bots that crawl the web for content and compose personalized articles that some might see as very good quality article for an AI to compose.
For example, the following diagram suggests modified version of this workflow:
The diagram above show cases a workflow where an article is completely created by the AI from start to submission for triage stage. In some other cases, a content writer might want to write the content and verify resources which has been returned by the AI crawler in this case, on his/her own. In this case, we could have some workflow similar to this:
Targeting and Personalization
AI Algorithms are capable of assisting publishers to target individual users based on their behavioral patterns and predict their needs, hence it can craft a personalized advertisement for individuals to boost customer engagement and satisfaction.
This article gave a glimpse of what it can be achieved if an AI got integrated into the core of a CMS. Furthermore, there are many use cases and other automation features that might make it easier to make quality contents and as a consequence serve the readers with useful relevant information.