Monday, April 19, 2010

The need for Metadata on Enterprise Knowledge Systems

OK, I admit that I use Google search a lot. I won't deny the importance of search in my life. These days, search is a convenient entry point into our complex intranets. However, I don't believe that layering search over document management systems is the answer to today's knowledge problem.

Here's why I think so:
  • Knowledge is linked. While search can identify physical connections between information, it cannot show qualitative relationships.
  • Knowledge is contextual. Search cannot identify the contextual significance of information.
  • Knowledge is valuable. Search by itself can't show value -- it needs 'intelligence' to decide value.
This is where I see the importance of metadata on knowledge systems. Metadata, coupled with search makes enterprise knowledge systems as useful as the web for finding the right information.

BTW, what is metadata?
"Metadata is a concept that applies mainly to electronically archived data and is used to describe the
  • definition;
  • structure;
  • administration
of data files..." - Wikipedia
In today's post I want to outline the power of metadata for your enterprise knowledge platform. Hopefully, you can use this as a guide to choose your knowledge platform if you're at that stage of your journey.

The Power to Relate Information

The ability to show relationships between information, is key for knowledge platforms. A lot of systems do this through tags. On learninggeneralist.com, you'll notice that there's a 'label list' on the website that relates articles about similar topics. If you've used social bookmarking, you'll notice that we categorise information under tags. On enterprise systems, collective tagging helps relationships and structure emerge. It then becomes easy for people to look at a certain tag, to reveal all information possibly connected to it. For example, I've tagged my favourite enterprise 2.0 case studies here. This holds great potential for induction, onboarding and capability building.

The Power to Show Value and Appropriateness

When everyone has the power to create useful information, everyone also has the power to create havoc through inappropriate contributions and misinformation. This is a risk that most execs are concerned about. Enterprise 2.0 systems mitigate this through 'more eyeballs looking'. We want users to have the power to say how valuable or suitable a contribution is. You must have seen several ways of doing this on the web.
  • Favourites - if a lot of people 'favourite' an article, then it's perhaps valuable.
  • Thumbs up or down - similar to favourites, except the thumb down can indicate if an article is unsuitable.
  • Flagging - allows you to flag items as inappropriate.
  • Rating scales - a quantitative way to suggest the value of specific content.

The Power to Contextualise Information

"Information in context, trumps information out of context." - Karl Kapp
Dr. Kapp's quote is a clear indication of why enterprise 2.0 systems are becoming the rage today. The ability to find useful, bite-sized information just when you need it is indeed the killer advantage. The ability to further contextualise this information makes it more relevant to the user's situation. On the web, we see this context emerge through comments. Take a look at blogs and associated comments for example. Free-form comments serve a number of purposes:
  • They allow users to show their reactions to the content.
  • They allow users to give feedback about the content. e.g. "I used this technique at work and ..." or "Here's another perspective..." or "I want to add that..."
  • They allow other users to use the content effectively. User commentary gives different perspectives and much needed context for the information in question.

This is just the tip of the iceberg when it comes to the benefits of metadata on enterprise knowledge systems. As you can see, the right kind of metadata can make your knowledge infrastructure come alive! I strongly believe that a knowledge platform based only on search is similar to putting lipstick on a pig! To make your knowledge come alive, you need users to contribute through low friction means. Comments, ratings, tagging create low barrier methods to make sure this happens. So do think about these elements when choosing your knowledge platform. Dinesh Tantri knows a lot more about these things than me, so do follow his blog here. Hope you enjoyed this post, please let me know what you think -- your comments are always valuable!

4 comments:

Rakesh Poddar said...

Quite an insightful post on enterprise content management, I must say.

In summary, you are saying that three types of metadata can enhance the chances of a user finding the right content - tags, ratings, and comments.

Out of these, ratings and comments have to be obviously generated by the consumers of content. What about tags? Can they be generated by consumers of content, or only by the producers of the content?

Thanks and regards,
Rakesh Poddar

Elena B. said...

Very good summary of what is needed basically to turn information into related knowledge: it turns out contextualization revolves around participation and some sort of conversation.

Concerning Rakesh question on tags, I think tags can be both bottom-up (by users) and top-bottom (defined by any CKO or admin), and tags can be added by both producers and consumers...and by adding tags, somehow the consumer of content stops being a 'consumer' to turn into a prosumer with their contribution on contextualization.

Cheers.

Sumeet Moghe said...

Thanks for your comments guys. And Elena, you're spot on about how we need to generate tags.

essay writing said...

Cool! Author the best! Thank you!

Related Posts with Thumbnails