Overview 1/99
for current version go to: http://www.wisdombase.org/filterinfo.html
In the last few years, the explosive growth of the internet has forever changed the way human beings communicate and access information. Yet this global meta-nervous system is still in its infancy. Today there are millions of pages of information on every imaginable subject available at the click of a mouse. The problem now is that there is simply too much information with no practical way to separate the wheat from the chaff.
This is where collaborative knowledge concentrating software comes in. We believe that collaborative information filtering and concentrating will be the next logical step for the World Wide Web, and we are developing a set of tools that will automate the collaborative construction of user friendly filtered knowledge-bases on virtually any subject:
Knowsys Knowledge Management Tools
FilterBase
- Large Scale Collaborate Groupware System
RealityCheck - Consensus Building Software
PeerReview - Collaborate Research Tools
FilterForum - Self-moderating Discussion Forum Software
What is Knowledge
Concentrating software?
Commercial and Practical
Applications
Background - The History of Knowledge
Filtering
Knowledge Beyond the Limits of Science
The KnowledgeFilter Window: Key Components
How Knowledge Filtering Works
The Answer Machine - The Query Routing
User Interface
Simple KnowledgeFilter Examples
The Wisdombase - Home
Page
Frequently asked Questions
What is Knowledge Concentrating software?
Knowledge Concentrating
software is a new filtering and sorting technology that allows
large groups of people to combine their knowledge and perspectives
into collectively generated, self-organizing knowledge-bases.
The software can either be used to create new knowledge structures
or to filter existing information for parameters such as percieved
veracity, effectiveness or other agreed upon measures of value.
It works like a large scale "word-of-mouth", allowing
users to compare notes with hundreds or thousands of other people
to find out what might be true or effective in virtually any area
of human knowledge. It allows individuals to participate in building,
filtering and accessing large scale knowledge-bases. The filtering
and sorting technology at the heart of KnowSys Software can be
used to:
The KnowledgeFilter software has three key
components. First, it allows users to contribute information,
opinions or links to information and to suggest or create new
sub-divisions within any topic. Second, it keeps track of how
others respond to and rate contributions or links. Third it sorts
the postings according to user feedback, allowing for sorting
by various criteria such as effectiveness, popularity or consistency
with information from other sources (number of corroborating links
or comments). It also allows the information to be sorted by source
or type of user (lay or professional in a given field, published
or unpublished material, controversial or widely held opinion).
Ultimately the software distills information and becomes what
amounts to a collectively generated knowledge-base with a built
in "truth filter". In its final form the software will
include a "query routing" system which will allow users
to access information in collectively generated knowledge-bases
by asking plain language questions.
Commercial and Practical Applications
Consensus building software will be invaluable for any group wanting to share information, access information, or build a collective body of knowledge on any particular subject. The filtering process allows large groups of experts to combine their knowledge and experience into a large-scale expert system. The user interface allows expert and lay users alike to access filtered knowledge by asking plain language questions or conducting key word searches. A topic-specific KnowledgeFilter would be a valuable asset to any website trying to build demographically focused traffic. It would also be invaluable for building practical knowledge-bases within organizations.
Our intention is to distribute the program as a software shell, much like an expert system shell with blank fields ready to customize to any subject area & sorting criteria. Knowledge-bases created by KnowledgeFilter software grow and develop organically and autonomously in response to user contributions, feedback, and democratically selected links. The only limit on the depth of insight that can be embedded into an autocybergenic knowledge-base is the available storage space, which is now virtually unlimited.
We believe that collective information filtering will be the next "killer application" for the rapidly evolving planetary nervous system that the internet has become.
Background
The History of Knowledge Filtering
Science itself has been called a knowledge filter. In the scientific model, individual ideas and discoveries are presented to the world or the scientific community for testing and scrutiny. The successful ideas become incorporated into a continually growing body of shared knowledge and are eventually transformed into useful things like airplanes and powerful medicines. Ideas that don't stand up to the scrutiny of humanly mediated reality testing fall be the wayside. Books are also crude knowledge filters. In the days before the printing press, only information that was considered valuable was hand transcribed for others to read. This might be considered the next generation of knowledge filtering after "word of mouth". With the invention of the printing press the filtering function fell under the influence of various economic and political forces, which did not always function in the service of truth. Despite these factors,the ideas that have reached into the present through the re-publishing of books are the ones that have stood the test of time because people found them valuable enough to pass from generation to generation.
Today information is filtered in a very different way. The information that reaches us is generally information that sells us on something, transferring wealth or power to the purveyor of the information. The two-way communication possible over the internet combined with automated information gathering and sorting technologies can rapidly change the quality of information we have at our disposal. As we enter the age of large scale dissemination of information it will become more and more important to develop mechanisms to "separate the wheat from the chaff".
Democratic Knowledge Filtering
Word of mouth has always been one of the best ways of getting at the truth. Every movie ad tries to convince us that we will love the film being marketed. Every product label tries to sell us on its product. But when several trusted friends tell us that we should see a movie or try a particular product we have more reason to believe them. Imagine what it would be like if we could quickly consult a collectively generated online resource that could tell us which products, methods or ideas have been most effective for the most people in our situation. Imagine if you could ask a question and then have the answers sorted in such a way that the criteria that you specify would cause the best answers to rise to the top of a long list of possibilities suggested by hundreds or thousands of people.
Knowledge Beyond the Limits of Science
Science has been an extremely successful way to filter information when it comes to dealing with the material world. Unfortunately some of the most important issues for us as human being are hard to approach with the tools of the scientific method. Science depends on objective measurement and the ability to reproduce results to determine objective truth. Yet many issues of infinite importance to us as human being involve things that are difficult to measure or reproduce. Collaborative Knowledge filtering allows groups of people to reliably formulate effective answers to subjective issues beyond the grasp of the scientific method.
Simple KnowledgeFilter Examples
Any commercial website would benefit from having a KnowledgeFilter relating to its specific area of interest. Public service organizations could use filtered knowledge-bases to gather and disseminate needed information. To begin to visualize how a simple filtered collaborate effort might look, consider the following:
A KnowledgeFilter cookbook
Users could contribute and access recipes or begin new topic "threads" for any type of culinary item. Feedback from users who tried the recipes would result in the best recipes rising to the top of each recipe topic list. Recipes could be sorted by criteria such as taste, nutrition and ease of preparation. If such a site were supported by enough people it could include thousands of recipes and variations from every culture with the best versions of each dish rising to the top of a long list of contributions. Users would come back again and again, commenting on recipes they tried and submitting recipes for consideration by other participants. (see sample cookbook window and cookbook index pages below )
A KnowledgeFilter joke center
A simple rating filter could allow users to view the "cream of the crop" of hundreds or thousands of jokes submitted by other users. New joke subject areas would be created by majority consensus just as new subjects are created in any other KnowledgeFilter application.
A KnowledgeFilter gardening advisor
People in a specific climate or region could use a FilterForum discussion group to compare notes and gather information pertinent to gardening in their particular area.
A KnowledgeFilter medical advisor
The software could be used to build a mega-knowledge-base for doctors and their patients. Postings could be sorted by expert opinion, number of corroborating links or references, or by personal experiences of patients and the lay public. A GroupMind version of the filtering software could be used for developing consensus and conducting collaborative research among professionals in any medical field. A simple web-based tool for a large scale comparing of notes among doctors and researchers could accelerate the collaborative discovery process now limited to peer reviewed medical and scientific journals.
KnowledgeFilter movie or book reviews and consumer reports
The RealityCheck version of the filtering software could easily be used to create online forums that would allow consumers of any product to get consensus feedback from large groups of other consumers.
Reputation Brokering
The RealityCheck version of the filtering software could also be used to allow people to share information on reliability of businesses, institutions, and individuals. The commercial implications for this type of application are obvious: Reputable companies and individual professionals would pay to have their reputation tracked by a system similar to "value-star. Wide spread use of a system like this would improve our business environment in many ways.
KnowledgeFilter research projects
The ability of the software to sift and filter information makes it useful for building knowledge-bases, solving problems or improving performance within an organization. Any group of people trying to develop a product or solve a problem could brainstorm using a FilterForum discussion forum. Contributions that people found useful or worth considering would rise to the top of the posting hierarchy, allowing them to be seen and considered by other members of a large special interest group. The result would be a sort of democracy of ideas leading to true group-mind problem solving.
Non- Profit applications of knowledge filtering software
WisdomBase.org -The ultimate filtered knowledge-base
The original intention for the development of the KnowledgeFilter software was as a tool to create a self-organizing interactive mega-encyclopedia of human self-knowledge and self-help technologies. Our vision is to create filtered knowledge-base or "wisdombase" that could become a collaboratively-generated, cross-cultural tribal elder for our new global village. Easy to use query routing software and collaboratively developed links to other information sources on the World Wide Web would make it user friendly and increasingly broad based. It is an endeavor to gather our collective knowledge about ourselves into a practical, friendly resource available to all. An "Online Help" system for our human software. The construction site for this project can be seen at www.wisdombase.org.
Our ultimate goal for the commercial development and dissemination of the KnowledgeFilter software is to finance and launch the wisdombase and to perfect the technology needed to realize the KnowSys vision.
The KnowledgeFilter Window: Key Components
The generative user interface will be in the form of an information window. A window can contain a posting from an individual, an existing piece of accepted knowledge, or a news group or BBS posting imported into the program. It can also contain be a window onto information or viewpoints contained at a specific spot on another website.
Users will be asked to provide feedback as they leave the page. Feedback options will include a quick-click assessment of perceived value, relevance, effectiveness, or other group chosen criteria.The software allows users to nominate and vote on the criteria used in building the knowledge-base. Mathematical values corresponding to user feedback statistics are invisibly assigned to each posting and end-users can sort the results by any combination of listed criteria. In the filtered cookbook example, a user may chose a custom sorting criteria that is 50% taste, 25% nutritional value and 25% ease of preparation for listing group contributed recipes for a certain dish.
Users can also indicate if they consider themselves experts on the subject or not. This information will be embedded in user ID's of contributors in the case of the more technically rigorous GroupMind research software . This information is then posted to a background look-up table where it is sorted and recalculated for inclusion in the "reality check" field of the posting window the next time it appears for any user.
The Sorted Index list Window: Key Components
There are two types of windows, content windows
and index windows. An index window will be a list of postings
within a specific heading and will appear on the initial screen
for that subject sorted by the mathematically derived "composite
truth rating" or "composite value rating". the
user can then select to re-sort the list by other criteria such
as:
click
here to go to an html version of this page without the comments
provided below
click
here to go to an html version of this page without the comments
provided below
The following chart shows how user contributions are filtered by feedback from other users and either become part of the main body of The Wisdom-Base, move to the searchable archives, or stimulate the creation of new discussion areas:
The following chart shows how end users will be able to access the knowledge-base by asking plain language questions:
for answers to questions like.......
Where does the information in a knowledge-base
come from?
What is the bias built into a system like this?
Aren't the masses really stupid... won't a knowledge democracy
backfire?
What will motivate people to contribute information, content or
votes?
Isn't truth a relative thing?
Doesn't software like this already exist?
GO TO: Frequently asked Questions
To visit the construction site for The KnowSys Project go to:
For more information on the KnowledgeFilter Software or The KnowSys Project please contact:
Michael Heumann
1717 Darby Rd. Sebastopol Ca. 95472
phone: (707) 829-0127
fax: (707) 829-9542
email: michael@wisdombase.org
BACK TO WISDOMBASE.ORG HOME PAGE