Best method to split an excessively large DT database

I’ve come to realize that the AI of DT is not performing as well as I would have liked, and I am realizing that the reason for this is probably entirely my own fault…I first decided to import pretty much every document on my computer into a single database, and the imported topics vary from one subject to another tremendously…

I wanted to address this issue by splitting up my database into a couple different databases, each of which would contain documents that pertain to a specific subject (at least relatively so…)

What is the best way to split up a database? Should I group all documents that I want to remove from the current database, then simply export this group to the finder, and subsequently proceed with creating an entirely new database and import the recently exported documents from the Finder into the new database?

Or is there a better solution?

Any suggestions would be greatly appreciated… Thanks!

The method you suggested, grouping the items to be exported, then importing them into a new database (and subsequently deleting them from the original database) is the method I use.

I don’t split databases merely by disciplines. My main database – reflecting interests in environmental science, law and policy – contains references and notes covering a lot of different disciplines, from chemistry to ecology to genetics to toxicology to economics, law and policy. What ties those materials together is that they all deal with my interest field. And because they do, I get a lot of useful suggestions from See Also. I especially value those that help me get insights into relationships between ideas that I hadn’t thought of.

In that context, I don’t want See Also to give me a list of references that are all about the same thing, say articles about bathtubs. I can do that with a search or two. What I want See Also to do is help me explore relationships between, say, bathtubs and something else that’s interesting. For example, if you look at the history of England you will find there’s an interesting relationship between bathtubs and the development of mass transportation, railroads. Bathtubs didn’t come into wide use until people started congregating together in closely packed quarters. (The British used to complain that the French concealed body odors by using perfumes, rather than bathing.) And of course neither bathtubs nor railroad tickets could have been afforded by many people until the economics of their society had undergone drastic changes. Still another thread of interest might be medical discoveries of factors that contribute to diseases, and growing interest in personal hygiene related to those discoveries – hence an interest in more frequent bathing, and all kinds of specially-designed baths and mineral waters became a fad. We shouldn’t neglect the development of urban systems to distribute water and to manage the resulting sewage, either. It all tied together.

Interacting with See Also suggestions and following trails of possibly interesting connections (not all of which may turn out to be really interesting) is what I like to do. DEVONthink’s ability to analyze contextual patterns of word use makes such interaction immensely more efficient than merely rummaging around in a collection of articles and books.