Discriminating Importance: Meaningful Organization of Search Results and the Democratization of Knowledge Management
Recent years have seen a proliferation of tools to organize and visualize various data sets (search results, lists, databases) into meaningful topic areas or hierarchies. For example, there's Grokker, an online tool which can reorder a simple linear list of search engine results into topic groups of probably-related pages. TouchGraph is a graphic browser for already-linked information such as Wikis or immediately-interpret-able links such as search results from the Google API. FreeMind, an open-source mind-mapping project, provides an extremely focused and simple interface for tree-oriented organization of data, and the formatting and linking of that data in meaningful ways. It also allows convenient import and export of data to standard formats.While metadata and XML are truly maturing in both specification and implementation, more "democratic" methods of classification and relation (folksonomy) are developing, most notably, the "tags" used at sites such as del.icio.us and flickr, where the highly individual, idiosyncratic and radically open system of tagging images or links with one's own system of meaning have congealed into surprisingly meaningful (to an audience of >1), and surprisingly self-organizing topic groups.
The Internet public, if such a beast can be defined, seems to be growing ever more comfortable with community software development (FireFox is a notable example) and community knowledge management (democratic or unmediated presentation of knowledge, of which the Wikipedia, and Wiki software in general, are perhaps the most dramatic examples). We are becoming participants more and more.
The surprise to me is the persistence of some unnecessary and non-productive dichotomies:
...and...
It seems obvious that these systems benefit from each other: the formalized standard learns what is truly needed & used from observing the larger community, while the larger community can save time and "stand on the shoulders of giants" by exploiting as needed proven and vetted expert and organizational knowledge.
It seems obvious that the formal and hierarchical serves well as the entry point and high-level structure, while the more democratic and dynamic forms (if they can be called that!) of knowledge management such as wikis, tags, and dynamic information mapping/clustering serve the searcher better at the chaotic edge of the formal classification, where the proliferation of singular items is so vast as to render static classification absurd.
It seems obvious that both machines and humans can contribute, both a priori and on-the-fly, to the meaningful organization of information. One simple example would be a system allowing for both the user and his machine agent (or chosen information portal) to deposit tags with appropriate "decay" rates allowing for the meaningful analysis and evaluation, by human or machine, of the relevance of a particular search path or segment. A kind of "search phosphorescence" if you will, from which intelligence could further manipulate and modify searches as they are still happening, forecasting relevance.
Federated searches exist, but we are still a long, long way from a kind of Unified Field Theory of search and research results--like sands through the hourglass, we sift, we sift, we sift...
The Internet public, if such a beast can be defined, seems to be growing ever more comfortable with community software development (FireFox is a notable example) and community knowledge management (democratic or unmediated presentation of knowledge, of which the Wikipedia, and Wiki software in general, are perhaps the most dramatic examples). We are becoming participants more and more.
The surprise to me is the persistence of some unnecessary and non-productive dichotomies:
- meritocracy vs. academic credentials
- meritocracy vs. participatory or consensus-driven democracy
...and...
- hierarchical & peer-reviewed vs. dynamically re-centering data-sets & rhizome-like, recursive structures,
- metadata and trees vs. tags and dynamic maps
It seems obvious that these systems benefit from each other: the formalized standard learns what is truly needed & used from observing the larger community, while the larger community can save time and "stand on the shoulders of giants" by exploiting as needed proven and vetted expert and organizational knowledge.
It seems obvious that the formal and hierarchical serves well as the entry point and high-level structure, while the more democratic and dynamic forms (if they can be called that!) of knowledge management such as wikis, tags, and dynamic information mapping/clustering serve the searcher better at the chaotic edge of the formal classification, where the proliferation of singular items is so vast as to render static classification absurd.
It seems obvious that both machines and humans can contribute, both a priori and on-the-fly, to the meaningful organization of information. One simple example would be a system allowing for both the user and his machine agent (or chosen information portal) to deposit tags with appropriate "decay" rates allowing for the meaningful analysis and evaluation, by human or machine, of the relevance of a particular search path or segment. A kind of "search phosphorescence" if you will, from which intelligence could further manipulate and modify searches as they are still happening, forecasting relevance.
Federated searches exist, but we are still a long, long way from a kind of Unified Field Theory of search and research results--like sands through the hourglass, we sift, we sift, we sift...
posted by Paul at 11:33
0 comments
