Comment What's needed is open-source search-engine (Score 1) 36
The world needs comprehensive, unbiased information about what resources are available on the web.
It is increasingly difficult to get "unbiased" results from commercial search engines. To my knowledge, none of the commercial search engines provide a complete description of how their engine works. The search algorithms change frequently. There is no requirement that the search engine provider disclose any financial relation it has with a referenced site. Since the internals of these search engines are hidden from public view, it is impossible to predict what types of results one should expect.
Which pages are included? Does the site fail to reference certain sites deemed objectionable? How are keywords determined? How are the results ordered?
What we can do, as those who realize the desirability and importance of the free and open exchange of ideas, is establish a freely-available database of all publicly available sites and a extensible, highly configurable search-engine based on open standards.
Unlike proprietary black-box search-engines, this search-engine should provide formally described services. It should be completely clear how the engine responds to queries. To ensure the integrity of the engine, the source code to the search engine as well as the formal descriptions of the search-engine services should be freely available.
Clients of the search-engine should be able to develop customized search strategies tailored to her particular needs, desires and biases. These search strategies can be shared with other users who may not wish to develop their own search strategies.
Many different front ends to the search engine may be created. The search-engine should be powerful enough to emulate the types of services provided by Yahoo, Infoseek, etc. Of course, a Yahoo emulator would place links to companies paying it money at the of its list. An Infoseek emulator would give special priority to offical ABC and Disney sites. Jerry Faldwell could create a specialized search engine that omits references to certain purple creatures seen on public television.
Meanwhile, the rest of us can have results free of hidden bias and targeted advertisements.
A group of netizens committed to the free and open exchange of ideas can establish a global, distributed network of machines to provide the aforementioned search services. The database provided by these servers should be as comprehensive as physically possible. All available sites should be indexed in the database. These sites should be committed to indexing all sites.
The privacy of users of search-engines should be respected. Therefore, providers of these public-servers should attest that they
will not collect information that links particular users to a set of queries.
So, is anyone working on a project like this?
michaelawells@mindspring.com
It is increasingly difficult to get "unbiased" results from commercial search engines. To my knowledge, none of the commercial search engines provide a complete description of how their engine works. The search algorithms change frequently. There is no requirement that the search engine provider disclose any financial relation it has with a referenced site. Since the internals of these search engines are hidden from public view, it is impossible to predict what types of results one should expect.
Which pages are included? Does the site fail to reference certain sites deemed objectionable? How are keywords determined? How are the results ordered?
What we can do, as those who realize the desirability and importance of the free and open exchange of ideas, is establish a freely-available database of all publicly available sites and a extensible, highly configurable search-engine based on open standards.
Unlike proprietary black-box search-engines, this search-engine should provide formally described services. It should be completely clear how the engine responds to queries. To ensure the integrity of the engine, the source code to the search engine as well as the formal descriptions of the search-engine services should be freely available.
Clients of the search-engine should be able to develop customized search strategies tailored to her particular needs, desires and biases. These search strategies can be shared with other users who may not wish to develop their own search strategies.
Many different front ends to the search engine may be created. The search-engine should be powerful enough to emulate the types of services provided by Yahoo, Infoseek, etc. Of course, a Yahoo emulator would place links to companies paying it money at the of its list. An Infoseek emulator would give special priority to offical ABC and Disney sites. Jerry Faldwell could create a specialized search engine that omits references to certain purple creatures seen on public television.
Meanwhile, the rest of us can have results free of hidden bias and targeted advertisements.
A group of netizens committed to the free and open exchange of ideas can establish a global, distributed network of machines to provide the aforementioned search services. The database provided by these servers should be as comprehensive as physically possible. All available sites should be indexed in the database. These sites should be committed to indexing all sites.
The privacy of users of search-engines should be respected. Therefore, providers of these public-servers should attest that they
will not collect information that links particular users to a set of queries.
So, is anyone working on a project like this?
michaelawells@mindspring.com
