For many people, doing a document search involves typing in a word and hoping for the best. For others, they have a specific document search strategy that they’ve honed over time.
There’s another group, and that is people who do some pre-work up front to be able to set themselves up for success when they need to find a document.
Awesome DocumentSnap reader Eric from Colorado is part of that group, and he wrote me to tell me about his unique search strategy. I haven’t heard anyone do things quite this way, so I thought I’d share it with the DocumentSnap audience. Take it away, Eric:
Group, Don’t Split
I do not want my search to find a specific document, I want the search to turn my mountain of documents into a small pile of ~ 20 documents, from which I pick out the one I want by eyeballing. I think of this as grouping rather than splitting. Or excluding with confidence 99% of my big pile so that I am left with the easy task of picking out the document I want from a small pile.
I came up with this approach a couple of years ago when my wife’s photo collection had grown to a size that she could not find anything, and her attempts to make albums and folders and descriptors had failed. I realized that a short list of keywords, in conjunction with the camera’s metadata of place and time was a strong enough filter. It boiled down to what, when, and where. When and where are automatic from the camera; for our purposes ‘what’ was broken down into the keywords things, views and faces. That’s it!
Limit Your Keywords
I apply the same idea to my document pile: I have ~ 10 keywords (no more!!) and usually attach 2 – 4 to a document. The trick is to pick keywords wisely that:
- Span the range of future documents.
- Spread the documents out evenly across the keywords; and
- Are VERY easy to remember.
Picking good keywords is an exercise in abstraction, which tends to be harder for people but is very powerful. In conjunction with a time interval, the search results do not become an unwieldy size. Evernote lets me add keywords to groups of documents at a time, so it takes very little time.
My math skills are rusty, but I think the group sizes about like this:
If I add seven documents a day then time lets me narrow down my big pile to 2000 documents per year, or less if I can be more time specific. If a document has two keywords out of a possible ten then C(10,2) or 10!/8!2! groupings are possible, which works out to 45 sub-piles. Three keywords (10!/7!3!) gives me 120 sub-piles.
Avoid the temptation to add keywords as time goes on, and adopt the mantra: group, don’t split!
A Document Search Strategy Example
I decided to show you a contrived example rather than my own pile so that I could start with a pretty Venn diagram culled from the internet:
(Image Credit: Oswego City School District Regents Exam Prep Center)
Keywords:
Animals
Aesthetics
Health
Economics
Wild
Example: I want to find long haired, black dogs with short tails.
I search for Animals+Aesthetics-Wild and find all 24 of the above records and perhaps cats with long ears and black Llamas too. Lions and alligators are wild and so are not found by the search A visual search of the records result lets me pick that one record of interest.
The reason I have the note in the first place though is because of my interest in animals and in aesthetics, and that is all I want my keywords and subsequent search to capture.
Thanks for this Eric. A very unique but powerful approach. I love the idea for photos too. Eric uses Evernote tags for this, but it could be done with Spotlight tags on the Mac or even PDF comments.