next up previous contents
Next: Hybrid-Search Up: Previous Work Previous: Unstructured Search

Associative ``Search''

Interestingly the first attempt to describe a personal information assistant described what would later lead to the large body of hypertext research [9]. In his essay, As We May Think, Vannevar Bush asked his readers to

Consider a future device for individual use, which is a sort of mechanized private file and library. It needs a name, and to coin one at random, ``memex'' will do. A memex is a device in which an individual stores all his books, records, and communications, and which is mechanized so that it may be consulted with exceeding speed and flexibility. It is an enlarged intimate supplement to his memory.
The memex Bush describes is a mechanical device for building associations between ``documents.'' Subsequently a number of efforts were made to produce such as system in the digital domain, most noteably, Douglas Engelbart's system AUGMENT/NLS (oN Line System)[14]. Although both Bush's and Engelbart's visions were for the inidividual human, Hypertext developed in a different direction. Disk space and bandwidth were once a scarce comodities, and the information systems built tended to have an emphasis on instituitonal knowledge representation. To this day, especially with the advent of the WWW, hypertext systems tend to be used to display one specific view that is produced by the hypertext creator. The associations created between objects generally satisfies the author's agenda, and not necessarily anything the reader could conceive.

So what exactly do hypertext systems do for us? Most systems allow for the creation of named ``links'' between documents. These links imply relationships between the documents that a user may traverse to get from one document to another. The claim is that associations would seem to be more valuable than mere indexing because humans think in terms of associations [9]. Hypertext systems therefore allow humans to explicitly describe relationships between documents that may not have been derived by a machine. In the case where the hypertext system is in fact entirely text (rather than hypermedia) a number of proposals [2] [4] devise methods for generating links on-the-fly.

However, while hypertext systems are good for browsing, on their own, they are usually ill equipped for searching purposes.

In summary:


next up previous contents
Next: Hybrid-Search Up: Previous Work Previous: Unstructured Search

Copyright 1998, Eytan Adar (eytan@alum.mit.edu)