We do quite a bit of consulting around metadata and taxonomy strategies for organizations, and one thing we find is always a topic of discussion is "How much metadata is too much?" The answer from us usually is: "The only metadata you should consider is information that is really relevant for searching, relating, or repurposing content."
Once the metadata strategy is determined, there is a fine line drawn to whittle down the actual number of metadata fields you'd like to have, vs. how many fields you think will actually be filled out. Filling out metadata can be a time-consuming task. 50 fields of information is most cases would be way too much! Some organizations will hire on additional staff to help fill out metadata, or police and control the quality of metadata entered by others.
While limiting down the number of metadata fields required for users to fill out is smart, what if there were an easy way to get more metadata information without more work?
I'll leave Text Mining Engines (used for metadata and content entity extraction) for another blog post, but I had an idea the other day. I was talking to an old college friend who regularly uses speech to text software to help write journal articles, and I thought "What if we could speak the metadata?" Speech to text software has improved quite a bit over the years, so I can't image that it would be that hard to integrate speech to text to help someone fill out key taxonomy terms for a piece of content or image. I know that I could describe 50 different properties of an image in just a matter of seconds where it could take many minutes to fill out the same 50 properties on a metadata form.
So what do you think? Worth exploring or waste of time? Let me know in the comments.
Posted at 06:15 pm by Ivan Mironchuk
Contact DPCI to learn more about our Web content management systems,
digital asset management solutions,
online collaboration software,
custom development,
multi-channel publishing systems, and
workflow management solutions.