Unstructured vs. Structured: Why Knowledge Must be AI-Ready

Posted by Stephanie Ventura on Oct 17, 2017 11:59:45 AM

knowledge into funnel graphic

Artificial intelligence is a powerful tool that businesses of any size need to adopt. A knowledge base powered by AI transforms the way company information is created, managed, and delivered. This enables teams to collaborate with fewer blockers and allows teams to focus on strategic planning without worrying about the inability to access critical knowledge to perform their duties.

Once your knowledge base is structured and “AI-ready,” the information you store within becomes contextually aware which benefits employees immeasurably.

The drawbacks of unstructured information

You might think that your knowledge base is pretty organized, but that doesn’t mean it’s properly structured for a machine to comprehend the information that lies within. While documentation on software systems or best practices for cold calls might exist, the content itself isn’t considered structured. Page breaks, bullet points, or comments indicate structure to humans, but not to machines.

This is problematic for employees who need information and expect to receive a succinct answer. Typically, employees go to HR or IT with questions because they don’t want to spend time sifting through documentation to find the knowledge they need.

When a knowledge base is unstructured, an employee might ask a question like “how much does my company contribute to my 401(k)?” and receive a one-size-fits-all kind of answer. But this is a question that typically has multiple answers.

For instance, let’s say your company only matches your 401(k) contribution after you’ve been employed for six months. If you’ve been employed for five months, the answer to the inquiry above will differ from someone who asks that same question but has been employed for two years.

Unstructured knowledge bases possess information, but are incapable of recognizing the context in which that information is helpful. For a knowledge base to offer relevant information to individual employees, the information within must be structured, or in other words, machine readable.

What is means to structure your knowledge

Knowledge that is AI-ready, or is easily digestible by machines, is annotated. On the surface, a structured document might look a lot like your unstructured documents, but behind the scenes, they’re quite different.

Annotation is a process of labeling or tagging sections of content within a document. Just like you may bold a line item in Google Docs to convey importance, you can indicate importance in an AI-powered knowledge base. You can label sections to indicate what type of employee a sections refers to, i.e., the Sales team versus the Customer Success team, or an employee who works in Boston versus an employee that works in the Dublin office.

Additionally, the way you create new knowledge with AI is changing, but it’s just like following a template. Let’s say you input a title into the header section. That title is now recognized as a header by AI simply because of where you wrote it. You’re able to tag sections of knowledge with keywords like “Customer Success workflow” or “product updates” so that when an employee requests or searches for information using those words, that section of content is given to them.

The importance of contextual awareness

Structured, AI-ready information is powerful because of the contextual awareness it possesses. AI-supported knowledge can recognize similarities in multiple documents within your knowledge base which helps employees receive relevant information when it's appropriate. When content is updated or added to by one department, AI can surface this information to other teams that interact with and possibly depend on that content.

Simply put, if someone asks, “can I change my tax deductions?” an intelligent agent can answer, “yes, at any time” and then go beyond and include the W-2 document, along with instructions about changing elections or calculating their new paycheck amount that live within the knowledge base, but weren’t specifically requested.

Context is also affected by the content creation date. For instance, if HR provided an explanation of last year’s health insurance benefits, that doesn’t help anyone if the benefits have changed this year. AI can recognize when knowledge is at risk for becoming irrelevant and can remind authors of the content to update or ensure the freshness of it.

In short, AI-ready knowledge provides employees with contextually relevant information when they request or search for it, or when AI deems relevant content helpful. AI-powered knowledge bases ensure employee alignment and improve collaboration and communication across your organization.

The most innovative teams are integrating AI into their processes. Keep up with the competition and try Talla for Knowledge Management today.

Request a Demo

Topics: AI, Artificial Intelligence, annotation, knowledge management