Here at Glentworth we take a lot of pride in understanding how information collections develop, evolve and emerge. This blog entry is about how we can draw lessons about how to curate our information from everyday life.
Over the years I have collected quite a library. Several thousand books, not all of which I have read. When I moved interstate a couple years ago, I placed my library away. In a rush, I packed up my collection according to efficiency of storage - Fit the most books into the least number of boxes.
Now, years later I have returned to my house and want to re-establish my library in a purpose built room. However, there is a problem: The order that the books get unpacked in does not match necessarily, the sequence in which I may need to access the volumes in the library...
In orther words:The dense interim storage method I selected is not conducive to random access information retrieval.
So, how should I order my library?
By “subject” to be sure, whatever that means. Since in many ways this is a single-user collection I am quite free to determine my own classification system. Of course a title for a nonfiction book gives some clue as to its subject matter, but that may not always reflect what knowledge I have drawn (or will draw) from the volume. If I have read the book, I may choose based on my memory to classify it in a way that diverges from the clues on the cover.
How do I get my books into a coherent arrangement?
- I have emptied the boxes onto my library shelves in no particular sequence
- Sort each shelf into clusters of like subject. Always leaving the books on the same shelf that I found them.
Over time, a pattern of similar clusters will emerge on different shelves, allowing me to identify a collection of clusters. This emergent order can inform how I continue to arrange the collection:
- Merge clusters into “superclusters” on their own shelves
- Divide superclusters back into more specialised subject clusters
- Repeat!
Over time a random collection develops some order, and over more time one can identify clusters that should be merged. The lessons from this process I think are obvious:
When managing information, take care to enable clusters to emerge. That is to say, provide space for sorting and resorting activities. In an organisational context this implies that it is important to establish the expectations of the business that the results of Knowledge and Information Management will come over time and yet not necessarily in quantum related to each discrete project.
Curate the data. There is no substitute for knowing what your collection contains. Clarify with your stakeholders that the practise of Information Management involves people who have or will need to acquire, intimate knowledge about a particular collection or domain.
Happy sorting!
