This may be a dumb noob question, in which case I apologize, but I’ve pored over the manual and repeatedly searched the forums, and I can’t find the answer. Does DEVONthink’s AI function across multiple databases in version 2? (I’m using Pro Office, if that makes any difference.)
Here’s a usage scenario to better explain what I’m talking about.
I write fiction, and I’d like to use DT to store writing fragments and ideas and snippets and possible titles and bits of dialogue and whatnot that come to me when and as they come to me. (This is one of the reasons I can’t wait for the iPhone app!) To that end, I created a ‘Writing’ database and made a group called ‘Fragments’ in it, in which I’ve put a bunch of these free-floating ideas I’d been storing elsewhere. I also made a group in the Writing database called ‘Technique’, in which I’ve put some neat technique ideas (articles, columns, snippets — you get the idea) that I’ve come across and want to hold onto. Needless to say, I’ve added additional tags to a bunch of these fragments and technique files.
I’m also a science buff, and I like to collect science articles and studies that interest me, so I created a ‘Science’ database, started populating it with science stuff I’ve collected, and of course I’ve started tagging that stuff too.
What I’d like to happen is for DT to make connections not only between different documents inside a single database (e.g. between different writing fragments that might be semantically related or have similar tag sets or what have you, or between a fragment I’m looking at and some writing technique which might be relevant) but between documents in multiple databases, e.g. between a story idea in my Writing database and some articles in my Science database that might be relevant… except that doesn’t seem to be happening. (To test it, I put an article on continuity and discontinuity in fiction in my Science database, and yet no connection appeared to a fragment in my writing database that contains the word ‘discontinuity’ a whole bunch of times. More to the point, I’m currently working on a story that involves time travel, and… well, you get the idea.)
Does the program work this way, or is the AI limited to functioning on documents inside a single database? Given that I can make global smart groups which draw from multiple databases (super-awesome feature, BTW!) it seems to me the AI ought to make connections across multiple databases, or what’s the point of ever using more than one database and thus losing out on one of DT’s most compelling, unique and powerful features, its AI?
On a related note, the lack of tagging is why I gave up on DT back in the 1.5 days not too long after I first bought it, and the addition of tagging is why I’ve come back. Tagging lets you fully describe a document and enables you to fluidly search and collect and classify all kinds of documents, whereas relying on a hierarchical folder structure, as previous version of the program did, is hopelessly constraining, not to mention just plain impossible in many cases. I do realize that using multiple databases is, to a degree, a hierarchical system of organization, but to me it just seems like the broadest sort of classification, one which makes it easier to find what I’m looking for when I’m looking for a specific top-level category of documents, like, say, scanned utility bills, or science articles. But when I want to explore connections and come up with new ideas, I don’t want that classification system to become a constraint that strangles creativity and gets in the way of making the sort of connections that are one of the very biggest and most important reasons I want to use DT in the first place!
Many, many thanks in advance for any clarifications and advice on this.