Method for Providing an Attribute Bounded Network of Computers

Abstract

An attribute bounded network of computers is created. Computers in the distributed computer network are identified by specific attributes (e.g., a geographically bounded region) and a server is used to distribute processing assignments (e.g., addresses of electronic documents to be indexed) based upon the identified attributes. A peer-to-peer computer network is also provided based upon geographically bounded regions, the peer-to-peer computer network can be used to share electronic documents. A virtual community can be created within a peer-to-peer computer network by identifying peer computer connections based upon associated attributes (e.g., a geographically bounded region). The attribute bounded network of computers provides indexes having fresher data by allowing spidering of electronic documents more often than can be done by a central server indexing site.

Description

RELATED APPLICATIONS[0001]This application is a continuation of U.S. application Ser. No. 10 / 044,291, filed Jan. 10, 2002, which claims the benefit of U.S. Provisional Application No. 60 / 265,024, filed on Jan. 30, 2001, U.S. Provisional Application No. 60 / 261,025, filed on Jan. 11, 2001, U.S. Provisional Application No. 60 / 267,508, filed on Feb. 8, 2001, U.S. Provisional Application No. 60 / 268,060, filed on Feb. 13, 2001 and U.S. Provisional Application No. 60 / 282,124, filed on Apr. 9, 2001. The entire teachings of the above applications are incorporated herein by reference.BACKGROUND OF THE INVENTION[0002]This invention relates generally to computer networks, and more specifically to providing an attribute bounded network of computers.[0003]Two of the major challenges facing the World Wide Web (“Web”) are the freshness of data (frequency of update) and depth (quality of coverage) of indexes on data. When a search engine spiders a Web site to update an index, the index is fresh at t...

AI Technical Summary

Benefits of technology

[0016]One example of an attribute is a geographically bounded region, the geographically bounded region can be automatically generated based upon the requesting computer's physical address (e.g., Boston, Mass.) or computer's network address (e.g., IP address). An attribute can also describe a topical boundary, such as interests, careers, or associations with specific brands. For example, a computer can be identified as being associated with “dancing”, “lawyers” or “Nike”. In these cases an attribute bounded request can be matched to electronic document addresses (e.g., URLs) on the server computer that are also associated with those attributes. An attribute bounded request can result in the same electronic document addresses being assigned to multiple computers in the computer network. This redundancy provides improved reliability and timeliness to the processed results. The system also has the ability to assign electronic document addresses outside of the attribute bounded specified. This is useful when certain electronic document addresses remain unassigned after a certain period of time.

[0044]While there are methods of sharing resources over many computers to search the Web, the method of geographically organizing the users has intrinsic benefits to address the problems outlined, and can provide a better organized method that currently exists for organizing a network of users with the intent of indexing web pages on the Internet. This method can be imposed on top of, or in parallel with, other methods of ordering as someone skilled in the art could do with the above information. This is not designed to address all applications, and may not apply to someone sharing resources for finding the cure for cancer, but has merit for applications where geography is a factor

Problems solved by technology

The prior art present many problems when attempting to provide a network of computers, specifically a network of computers for distributed processing.

Pre-distributed computing models present problems in that the user of the client computer has no control over what pages his computer spiders.

In fact, a client computer could end up spidering pages of Web sites that are against the users direct beliefs, values, or affiliation.

For example, a vegetarian may not want his client computer to be used to index pages on preparing meat products.

SETI@home also suffers from problems in allocating distributed computing resources, it does not have the capability to assign blocks based upon any attributes defined by the client computers.

Existing pure peer-to-peer systems, such as Gnutella, do not provide a mechanism to limit the breath (i.e., the set of computers to contact at a specific level) or the depth (i.e., the number of levels) based upon an attribute bounded region.

Images