Just to share an unorthodox idea for a new kind of smart folder by following the thought of the very old but unforgettable Lotus Agenda methodology:
DTPO has two main methods of “smart” classifications. The AI and the smart group. AI accepts the dynamic decision of user, but the rules (groups are suggested by DTPO) of classification is not changeable by the users. Smart folder can tailor almost any rules for classification, but the rules setting is an “upfront” (user need to set the criteria before filing/classification) or “after” (if user need to include some other relevant files, he needs to set up new criteria) process. So, smart group is flexible but not dynamic in rules setting, AI is dynamic but not flexible in rules setting.
What makes lotus Agenda an unforgettable application is that its rules settings for classification are both dynamic and flexible. Applying the concept to a new type of smarter smart group:
A normal group is a group that we move/replicate/duplicate files because we think the files are relevant to a concept, a project, or a subject. It’s 100% based on users’ interpretation.
(1) function 1: what if we review an article and identify a new term, key word or some phrase that makes us to think that all articles with such key word should be assigned to a specific group. What if all we need to do is to highlight the key word/ phrase and right click to select the group we think the file should be assigned to. AND, the group (while accepting subjective file assignment from use under normal situation) will automatically include all files with the same key words? In that sense, group become a hybrid of smart and normal group, users can assign the file by their own decisions, or the categorization can be executed according to the new and dynamically set rules. The classification of files-groups becoming both dynamic (because users can set a new rules anytime) and flexible (because both soft/subjective user decision and hard/automatic rules can be used for classification in the same group).
This function will be a godsend when users have a bunch of files in certain parent group but want to categorise them into better defined child group. More importantly, the files can now be concurrently and automatically attached to different groups as long as they satisfy the rules in each group. This function will be another godsend when users need to do a lot of cross-reference analysis (project-nature, topic-concept, case-client, etc). Imagine a user just thought of a new key word relating to an idea during the note-taking process, all he needs is to set the rules on the spot. He can continue his writing/reading while all of his previous notes will be attached to right group immediately.
(2) function 2: chaos control. One lesson learnt by Agenda’s users was that when rules of categorisation become so dynamic and flexible, categorisation is out of control. Many files may suddenly appear in different groups automatically. There are two control methods. First, the user can simply delete the unwanted files (even they contain the key words) and DTPO will flag the files and never put them back into the group automatically - unless user change their mind and put them back subjectively. Second control method is to limit the source of files that can be automatically allocated to the group. The user can set up the limit once he discover that the categorization is out of control for a certain group - and all non-related files will be disappeared.
(3) Since the key words are now attached as rules to the group, the existing AI will become much smarter because the systems now have additional and reliable user preference as additional parameters to construct the algorthrims of auto classification suggestions.
I am a very happy DTPO user and are sure that the 80% of the functions I mentioned above can be achieved by the current DTPO settings, just in different ways.
I apologise in advance for such a crazy idea.