Included in my annual to do list is a review of my keywords. Those words I use to annotate the pictures in my catalogue. There are several objectives of the review:
- Validate the keywords continue to be useful
- Remove duplicates
- Merge the similar and overlapping
- Restructure as necessary
I need to put keywording in context. While I appreciate the value that keywording offers, it can represent a fair commitment of time and effort and to get it done properly (i.e., correctly, completely, and consistently) at least for this human mind, is nearly impossible. Long term, I expect this problem to go away with automation. Trends already point to this eventuality with face recognition capabilities built into software as well as automatic tagging capabilities emerging on sites like Flickr.
However, in the short term, the effort remains necessary. So I depend on a combined approach of a catalogue structure and a conceptually simple framework coupled with only keywording what is absolutely necessary.
The fundamental unit of grouping pictures is an event; all pictures I take are placed into one event.
I then establish a hierarchical catalogue structure by year and within each year are two sub-categories of events: [1] A Day in the Life [2] Travel.
A Day in the Life — representing daily events — is further subdivided into months and then by day. A folder (i.e., a Lightroom Collection) is created for an event, named with the name of the event and prefixed of the day of the month. So, for example, on December 13th, 2015 I attended a fashion shoot. That event is recorded by name “13 Fashion Shoot …” in the December folder of 2015.
The second folder, Travel — representing travel events– is organized simply by the trip name. I prefix the trip name with the month number the travel starts, just to keep them in sequence. Within each trip are “Smart Catalogues” that organized pictures by their rating and possibly a keyword. So, for example, in June we went to Poland so there is a folder (Lightroom Collection) names “06 Poland.”
Typically keywording travel images is the most onerous due to the volume; there could be hundreds or even thousands of pictures. In this case, I’ll complete baseline keywording to cover off the things that are important to me: all images get tagged with Event-Travel. Next images are tagged by image type (e.g. landscape); gps location (if not captured automatically); image location name (e.g., name of a park that GPS would not otherwise capture);
My next step is to rank the images with 0 to 4 stars. Following ranking I will then complete a second and more detailed keywording pass on those pictures that have 1 or more stars; it depends on how many there are.
I organized my keywords into an extensible framework according to their purpose:
– Descriptive of the image (Who, What, When, Where)
– Who is in the image
– What activity is being captured
– What event is being captured
– What is the type of image
– What are the things in the image
– When was the image taken
– Where was the image taken
– Descriptive of technical elements / construction (How)
– What compositional features are in the image
– What equipment was used
– Image Context (Why)
– What is the broader context (purpose) of the image
– Where the image is used
– Which books, websites, videos, printed, etc.
– Control on Image use
– Has it been registered; is it for public or family use only
As suggested above, I don’t necessarily code all of these dimensions for every picture; they are there just if I need them.
Having built on this framework over the last few years this is what I’ve learnt:
- Be clear on purpose:
Is keywording for personal or commercial use? If the former there is more flexibility with the choice of terms; for commercial use one needs to factor in commonly used terms and align with industry standards, consider SEO, etc. Extensive keywording is important for stock photography. I focus here on the personal side. - It’s iterative:
You can’t come up with a list of keywords in one shot, even if you buy a starter set; the list of keywords needs to be inclusive and extensible / open ended. It takes time to come up with clearly defined and distinct words - It’s personal:
the keywords need to be consistent with your mental model and needs - It can’t be heavy:
keywords should be optional; used only when necessary - Leverage automation:
Digital cameras can cover some of the requirements; automation with scene and face recognition is emerging. But you still need to know how to structure the words for coverage - I rarely use keywords to find images; I usually search by the Event
- My most frequently used keyword searches are:
– “Unregistered”: look for photographs that have not yet been registered for copyright
– Image type
– Camera Type
– People
Leave a Reply