Document Organisation Part 2 – Management
Overview
Similar to my photo organisation post I have also looked into doing something similar for documents. The goal is become paperless or at least less reliant on paper documentation. When I embarked on this workflow optimisation I wanted to make sure the paper documents I have whether it was till receipts to financial statements I would have them in electronic form so I could access them without having to find the physical copy.
Storage
Similar to my photo organisation post I wanted to use the file system to organise the documents and then use any programs on top. This is to keep the organisation software agnostic and doesn’t tie me into a specific program. The structure was designed to separate duplicate named documents and also to facilitate browsing by date.
As a general rule I prefer the Japanese date format because it makes it sortable in the file system. I use ~\[Year]\[Month] in 4 and 2 digits respectively for folders. Unlike photos I don’t see the go down to days because unlike photos I don’t envisage having a mass of documents being created or received.
The files are named in this format [Year]-[Month]-[Day] (+tag) – [file name]. An example of this is 2015-07-26 +certificate +education – Introduction to XML course.pdf. You can see the date is fairly self explanatory. Tags are one or more keywords with a plus (+) infront. This should allow searches to be performed using tags. The hyphen “-” separator is free form text to describe the document. This usually is text from the document (with some exceptions. See below File Naming section for more details).
Using the example above the path would be something like this:
2015\07\2015-07-26 +certificate +education - Introduction to XML course.pdf
Retrieving Documents
The storage method above makes browsing for documents by date very easily so long as you have the year and month. If you have a day then it will narrow down to the file within the folder structure.
The explorer search in Windows 8.1 also worked very well. Starting at the top directory you can type in a search criteria and it will search all sub folders highlighting the word it found. This makes tagging and naming your files very important.
This is no replacement for programs that can index files however this basic foundation gives you a lot for almost no additional cost. See here for more details.
Windows Automation
This doesn’t cover how to get the document to an electronic format. See here for more details.
My automation tool for Windows of choice is Dropit. Unfortunately it doesn’t support chaining very well. I use a profile to rename the file to a template file name. This takes the modified date of the file and put that date in first. It adds a plus sign ready for at least 1 tag and then the original file name. For example the file was named abc.pdf will become 2015-09-28 + – abc.pdf.
The next profile will check the file name has a date format at the beginning. I’d like to check that there is at least one tag after as a minimum but I haven’t managed to get it to work. Using the date format it moves the file to the relevant folder.
File Naming
There have been a lot of lessons learnt when I was going through this especially on file naming alone. Of course it’s continuously improving so these are not the final points that I have found to work. These also may not work for you.
- Keep tags short – the file names can be quite long themselves so the last thing you want to do is expand that by using long named tags. File names also have a limit.
- Remove special characters – Older file systems have problems with special characters in file names. Also consider if you were to use a script in the future how it will cope. Generally I replace them with a space.
- Consistency – Being consist in tags and names will help when you’re trying to find a bunch of documents of a certain type. Whilst it’s not always possible especially continuously improving on this as long as you’ve been consistent you can always go back and batch change them in a single pass rather than doing it several times because you’ve used different tag names to describe the same thing for example.
- Put company names in the file name – If they’re not present add the company name in the file name. Usually a company request information from you or you remember the company but not when the document was received. This helps with searching. An idea I had toyed with was using tags however some company names conflict with tags I’m already using. For example a company with the name “Enterprise Software” however I might be using that to classify my enterprise software documents.
Summary
It’s not the most elegant way of doing it but gets 80% done and quickly. The tagging is the most difficult part to keep them as consistent as possible. Also the key is the name of the file. The windows search feature is powerful but does not go as far as a dedicated document management system.
How to Speed Up, Disable, or Rebuild the Windows Search Index
