Having Fun With Semantic Metadata

Last week, I hosted a webinar on Semantic Metadata for the company previously known as Nstein (now Open Text) -- along with two of my colleagues, Sheila Woo and Matt Mullen. Despite putting it together in just under 10 days we still had over 250 registrants, which proves my point that interest in Semantic Metadata has created a "hockey stick" graph; slight incline for 10 years -- and soaring upward now.

Matt and Sheila are uniquely well-versed in how semantic metadata can assist the largest of information providers, so it was an engaging webcast. We looked at how Semantic Metadata can boost Audience Engagement, how it can increase traffic and stickiness, how we could improve productivity and contain costs -- and how it can provide a unique search tool on both the back end and the front. When we polled the audience, more than 40% said they could imagine using Semantic Metadata to improve audience engagement. This tracks with Gilbane's analysis that Audience Engagement is the number one focus of enterprise Marketers.

When we turned it loose for questions we must have piqued the imaginations of many -- because over 25 questions flew in. The range was from how long does it take to implement (the obvious answer is the ever-annoying "it depends) to how do I create a taxonomy (this is Sheila's forté - so she riffed on that) you can go generic with an IPTC taxonomy or do an analysis of all your assets to create a custom taxonomy.

People were particularly intrigued with the idea of putting a "linguistic fingerprint" on assets --using a combination of linguistic and computational analysis -- so you can find content that highly similar to another. Or, as we say, "relevant content."

We had someone ask how you could use Semantic Metadata to improve information security, that was my baby: consider adding it into an email process to determine if sensitive information is being sent out through the firewall.

The truth is, Semantic Metadata, which is the data that describes in rich detail the "aboutness" of something, can be incorporated into virtually any application that incorporates "unstructured" data. If you are a content freak like I am -- it's time to consider becoming a semantic metadata geek.