Most web pages are just pieces of the puzzle. Without the "Big Picture", it is hard to find the right piece at the right time. zKnowMan captures the temporal relationship between web pages surfed. Using user feedback and sophisticated algorithms, zKnowMan can analyze which documents are closer to what the user wants to find. zKnowMan also provides sophisticated analysis using Knowledge Management techniques such as standard taxonomies based on XML, such as OASIS.
zKnowMan automatically records all the web pages that a user goes to. The user can recall any web page by recalling any of the text on the page. zKnowMan allows full text searching of everything stored in it's cache. zKnowMan can optionally cache and index images and other non-HTML files such as Microsoft Word and PDF documents. zKnowMan uses cutting edge compression that allows the complete cache and index to be approximately 40% of the original size of the documents.
zKnowMan is designed to allow searching of multiple user's caches, effectively functioning as a distributed, peer to peer (P2P) web search engine. zKnowMan is intended to operate similarly to a search engine, such as google.com, in respect to DMCA compliance (specifically DMCA Section 512 "safe harbors" ) . It was designed based on insights from Fred von Lohmann's EFF paper.
zKnowMan is designed to be integratable with various P2P technologies such as Jabber and OpenCola using XML.
Accuracy in program performance comes from using predictable technolgies. An underlying principle is to use predictable and automated technology (Non-Turing) such as HTML, SQL, and state machine techniques like regular expressions as much as possible.
zKnowMan can run from a user's laptop or from a web server on the Internet. It is designed to allow both data replication and data portability. The cache is platform independent, and the text data can be exported as plain ASCII.
zKnowMan can be used from any web enabled device - from PC to Palms to wireless devices such as Blackberry's.
zKnowMan can be run on just about any computer with a C compiler - from Macs and PC's running Windows or Linux to Unix servers to IBM mainframes.
zKnowMan is designed to be accessible to all levels of users, from web surfers, to Javascript / Web designers, to programmers. All functionality can be access through XML-RPC.
zKnowMan represents the multi-dimensional aspects of information as layers on a map. zKnowMan divides the information into three main categories - Internal Information, External Information, and Processes that interact with the information.
Different maps of the same geographic area can show different, but related information. A Travel Map focuses on highways and streets. An Aviation Map focuses on flight related information - no fly zones, magnetic anomolies, etc. zKnowMan allows the user to define the focus, and display relevant information.
zKnowMan's foundational assumption is that relationships and analysis is best based on known, human verified information which can be automated. Examples are Netscape Open Directory (dmoz.org), Govt. databases (SIC, Census information, Library of Congress), and publically available information - (Phone books, industry standards).
For instance, given a companies' phone number, a user could find the companies' web site, which industry the company is in, govt. statistics on growth, industry magazines and trade organizations, as well as potential competitors. Another example is using the citation index - If you find a good article, who else makes references to it? Still another example is word relationships - like what is a more general term for poet - a writer.
The various layers of a map represent different databases. For the previous company example, there are a yellow pages layer, SIC layer, magazine layer, trade association layer, etc. zKnowMan scans the text of web pages for relavent terms in both the internal and external knowledge bases.
Most web pages are majority text content. HTML compression on the fly reduces data transfer up to 75%.
Increase productivity by removing unecessary distrationtions. Ad removal - and content filtering - help the user focus on the task at hand.
Security starts with SSL (Secure Sockets Layer), and includes a finely tunable security framework with ACL's and LDAP compatibility.
If a site is interesting, why manually click on all the pages? Automatic navigation and caching of an entire site can be automated using the W3C's libwww robot.
Use a modern, DHTML interface that leverages the best in Human Factors, such as Raskin's quasimodes.