September 4, 2014

Is Intelligent Content a Good Idea in Technical Communication?

Over the summer, I became a bit provoked while reading the thoughts and opinions of people in our industry regarding how technical communication is supposed to look in future. My overall impression left me mulling over the following question: Should technical communicators be designing content so that the right information presents itself to the right person, at the right time, on the right device, and in the right location?

Is it possible? Is this what users want? Here, in part 7 of the Designing for the Searching User series, I delve into the latest fixation in the world of technical communication— something I refer to here as Intelligent Content.

Read further to discover why I say the answer is NO— why I believe that any attempt to design intelligent content in the context of our industry would not only fail, but users would ultimately find such an approach useless. It is important to reflect upon this issue, since what we aim for in the future, informs the decisions we make today.

What is Intelligent Content All About?

The primary objective of technical communicators is to assist users through the process of completing a specific task. We design information to support users since we believe that once they have obtained that information, they will possess the knowledge they need to bring the task to completion. This objective is quite simple, and easily agreed upon between technical communicators.

However, the question of how information should be delivered to the user— and how they prefer to interact with it— is more complex, and remains a source of much debate.

As the debate continues, content-heavy organizations feel pressed to keep up the pace; constantly evolving and optimizing the way they create, manage, and deliver content. Some in our industry say that the ultimate goal of such optimization is delivering the right content to the right audience. Anywhere. Anytime. On any device.

What does this mean? Imagine a user who gets stuck while using a given software. In need of information, the user consults their intelligent content application, (or virtual assistant). Once it is opened, the content that solves the user’s information need automatically pops up. This, before the user has even expressed their query.

More and more, I hear people in our industry saying that designing intelligent content is possible since we have a variety of tools available that filter information. It’s claimed that intelligent content can be adapted automatically to each user, taking advantage of specific information we collect about them in order to deliver optimized content. All this happens on the web where content is seen as a utility.

How Does Intelligent Content Depict the User?

Personally, I smell Shannon & Weaver’s communication model from the late 1940s when I think of Intelligent Content. The sender codes a message, and then sends it to the receiver who decodes it. The receiver is passive, and cannot decode the message until the sender has coded and sent it.

Likewise, the Intelligent Content approach portrays the individual for whom the content is prepared as a passive creature, who is sitting and waiting for the technical communicator to deliver the right content, on the right device, at the right time.

The only one who is active here is the Intelligent Content application, developed by a technical communicator, which must filter and adapt content in order to customize it for the individual. If the passive individual should find the content they receive to be inappropriate, the application must back-track and re-filter it in order to get it right.

Let me present an analogy to paint a clearer picture. It begins with a hunter, who wants to strike down a prey. To do so, the hunter needs to carefully select a riffle and bullets suitable for the specific prey they are targeting. Once the hunter spots their prey, a well-aimed shot is delivered to strike it down. The prey, unaware of the hunters' activity, is dead before noticing anything.

The hunter in this analogy represents the technical communicator, who wants to deliver the right content to the right audience. The user becomes aware of being “hit," and their information need is "struck down” almost at the same time they perceive the need for information.

A problem for the technical communicator is that he or she must "shoot" before having their prey in sight; since there’s a time gap between the development of content, and the content adaptation that happens when the user perceives an information need.

Does this Depiction Correspond to What We Know About User Behavior?

Armies of information scientists have investigated information behavior for more than half a century; ( just explore the archives of Google Scholar on this topic). The findings from this research provide a good starting point to understanding how users behave and what they actually want.

Studies show that an individual may perceive an information need due to uncertainty which arises from an anomalous state of knowledge. If motivated, this individual may decide to initiate an information-seeking task to find the answers they need and reduce their uncertainty.

The information-seeker is active and goal oriented— selecting the most relevant and preferred information source, expressing inquires, and making relevant judgments in their pursuit to find the most relevant information.

Information-seeking behavior is complex, and at the heart of this complexity is the concept of relevance. Scientists struggle to understand how to design an information retrieval system that knows what information is relevant to a given query. How does an individual judge the relevance of an information source, and how do they judge the relevance of the information they find within it?

This research portrays the user quite differently than the Intelligent Content approach does. Rather than being a prey, the user is the active hunter, following an information scent. The technical communicator, (who is not the hunter), must design content so that it smells good to the information forager, rather than trying to shoot content at the user.

What's the Problem with Pacifying the User?

It is tempting to think that users really do want to get the right information, at the right time, on the right device, and in the right location. The user could carry around a virtual assistance device, like the artificially intelligent computer HAL 9000 in Stanley Kubrick’s “Space Odyssey,” or like SIRI on an iPhone.

But in my opinion, the content creation tools and knowledge we have today, and in the coming 50 years, make it very hard to design such assistance in the technical communication industry. Why?

Because users who have an information need, do not know what information is appropriate for their current concern— and they have even more trouble expressing what they need to someone else. If the user cannot say what they need, how can someone from the outside deliver the right content?

The problem is that a technical communicator, assured that they can deliver the right content, at the right time, on the right device, and in the right location, will most likely neglect to provide the user with the right “stuff”: a context, meaning metadata, classification and links to relevant follow up questions, etc. This “stuff” accompanying the actual information, is, from this viewpoint, unnecessary; especially since there is a cost involved in creating and maintaining it.

But this is the “stuff” that helps the searching user judge the relevance of the information they are receiving. As users make selections in TOCs, filter through data, express queries, and follow links, they build a mental model which is used as a frame of reference for understanding what information they actually need; and which ultimately helps them to judge, reflect, and understand what they read— which helps them to obtain new knowledge.

So the act of seeking proves to be important in synthesizing what is read. To display content for a user without providing any context, and without any seeking having taken place, annoys the user and results in a poor user experience.

What Should We Do Instead of Trying to Deliver the Right Content?

Instead of making content “intelligent,” we should focus on establishing a framework for understanding what type of questions users ask, and how they can be predicted. Furthermore, answers must be easy to find.

Instead of simply trying to "deliver the right information to the right audience,” we should aim to help the user judge the relevance of the information retrieved as a result of their inquiry. The challenge then becomes to know how to help the user in assessing which result is most relevant to their current information need.

Here, filters can be designed from the Situation Assessment Dialog, providing a means by which the user can filter out everything that is not relevant, while at the same time helping them to gather a broad sense of what the remaining results consist of.

My work on SeSAM aims at providing technical communicators with a framework for both predicting user questions, and designing searching user interfaces that help the user judge the relevance of information. For more on this topic, read other articles from the Designing for the Searching User series, and be sure to stay tuned for future additions.

Previous articles in the Designing for the Searching User series: Part 1Part 2Part 3Part 4/Part 5/Part 6

About the author

Jonatan Lundin

Jonatan is a pioneering information architect backed by over 20 years dedicated to XML documentation, and designing for findability.

Post Comment


  • Excosoft
  • Info Tech Trends
  • User Experience

Latest posts

More from Blog & News