Keywording Again: A Quick Look at Image Keyworder

After putting together Keywording for Fun and Profit I started looking around a bit for tools or services that would help me keyword large numbers of images, and came across Image Keyworder, and have been playing with it for the last day or two. This post will reflect my first impressions of the product.

Downloading and installation were simple and unproblematic, Image Keyworder (IK) runs only on Windows platforms and requires the installation of the Dot Net Framework 2.0 and SQL Server Express 2005, both of which are performed automatically for you by the installation process. Once IK is started and you register for the demo, it downloads the English thesaurus, you can then browse to files and start entering metadata. IK provides a set of tabs which allow you to review and edit the range of metadata you might wish to enter, but the core of the product is of course keywords and IK’s thesaurus, and so I’ll focus most of my attention on that part of the interface.

The intended workflow seems to start with searching for a particular word/idea (or a prefix of it), say “bear”. This will give you a list of matches from the thesaurus, from “Actions -> Bearing” to “Animals -> Land Mammals -> Bears”. Note that the thesaurus is structured, it knows that a “Polar Bear” is a “Bear” is a “Land Mammal” is a “Animal”. From there, I can move to the different parts of the “keyword tree” and have keywords added, for example, I might select “Polar Bear”, and if I hit “Add Keywords” after selecting that I’ll have a variety of keywords generated from “Bear” and “Mammal” to “Ursus maritimus” and “Polar Bear”, it’s easy then for me to cull out the ones that I think may not be on-target.

IK has no trouble inserting keywords into a JPEGs or a single-layer TIFF files, but does not support other image formats, most notably raw camera images. There’s no mechanism like the ones in Lightroom to allow you to manipulate the keyword lists of multiple images at the same time directly, there are however tools to allow cutting and pasting of keyword lists.

The thesaurus is mostly pretty strong, in working with a few images I’d hand-generated keywords for recently, I was able to come up with additional keywords that hadn’t occurred to me. Unlike a traditional book thesaurus IK gives appropriate weight to at least some important proper nouns, under Mountain Ranges you’ll find a list of some of the better known worldwide mountain ranges, under Bears you’ll find a list of types of Bear, this product is well tuned for the photographic task. Databases like this can never be complete, but it does have a broad selection of words that cover the needs of the stock photographer fairly well.

I’ve had a couple program crashes with IK, a UI lockup and at least once had to reregister for the demo. This product, like Lightroom, needs to spend more in QA next time around.

Image Keyworder is a product of OnAsia, can be downloaded from the product web site as a 30-day free demo, and sells for $79.99, plus a $39.99/year fee for updates to their thesaurus (the first year is included.) On the whole, it seems a useful product, and one I hope OnAsia continues to develop. With raw support and a little more time in QA this would be a killer product, as it is, it’s still quite useful. Check it out.