New approaches to information management: attribute-centric data systems

Trying to find information on the Web is like trying to find something at a jumble sale: it is fun, and you can make serendipitous discoveries, but for directed search it is better to go to a department store; there, someone has already done most of the arranging for you. Unfortunately, the Web's continuing explosion in size, its enormous diversity of topics, and its great volatility, make unaided human indexing impossible. This problem is just a special case of the general problem of organizing information to create knowledge. A similar problem arises on the desktop when dealing with file systems, where users must search by name. When searching for a particular file, however, users often do not remember the file's name or location. File names are artifacts of current operating systems, but human understanding neither requires objects to be named, nor does it have problems with multiple objects sharing properties, names, for instance. That more general approach is not developed in current file systems or user interfaces. We argue for an approach to information representation based on the use of attributes and search. This representation is organization-neutral, thereby giving a flexible substrate for anyone to build multiple simultaneous organizations. We argue the approach from three perspectives: Attribute Value System (AVS), a networked storage system where objects are composed solely of attribute-value pairs; DomainView (DV), a desktop metaphor where objects do not have explicit names and retrieval is done by content; and KnownSpace (KS), a personalized desktop data manager.