STEINKRUG UNVEILS GENEPAGE WEB ACTIVITY TECHNOLOGY

Steinkrug Publications has taken the wraps off a software technology that monitors a user's activity on the Web and creates a Web page file that encapsulates their research on the Web using artificial intelligence (AI) technology.

Known as GenePage, the software requires a copy of Internet Explorer and runs under Windows 95. According to Peter Kruger, the founder of Steinkrug Publications, the software is free to use on a non-commercial basis. Business or commercial users, meanwhile, pay a licensing fee of $49 per user, with volume discounts available for multiple user sites.

Kruger told Newsbytes that the original aim of GenePage was to allow, for example, an electronic researcher to have access to a fellow researcher's Web research history. The facility, he said, is particularly important in a library environment, where someone leaves a project and others have to pick up where that person left off.

"Once we'd developed the software we realized it could also be used by fellow researchers on a project to keep an eye on what the others were up to. It also allows employers to monitor what their staff are up to on the Web," he explained.

In use, Newsbytes notes, GenePage gathers and grades information describing a PC user's working environment and matches it with that of other Web users in a workgroup or on a company's intranet.

According to Kruger, by keeping the strongest data and discarding the weakest, GenePage provides an evolving collection of information which behaves like an extension of the human memory and thought process. Kruger said that the ability to assemble and grade information -- i.e. keeping strong data and discarding the weak -- enables the package to create an evolving networked database of resources.

In operation, he said, GenePage mimics, and acts as, an extension to human memory and thought processes. "It also blends the way we work in groups with the operation of an Internet enabled PC," he claimed.

According to Kruger, the software is based on a technology known as genetic algorithms (GAs), which are seen as the key to building a thinking computer, which is defined as a system that can develop and use thought processes similar to our own.

Kruger said that, while GAs are finding their way into software for toys and computer games practical applications in the workplace are rare. GenePage, he noted, is one such application.

GenePage consists of two main components; an editor to collect data to build up a profile of the user then publish it in the form of a HTML (hypertext markup language) page on the Web, and a browser which can be used to access, and merge with their own page, the GenePage files of other Web users.

According to Kruger, data used by GenePage to construct a page consists of keywords and URLs (uniform resource locators) derived from the user's Web accesses.

Each piece of information is value ranked by assigning to it a strength based on the number of times the user has accessed a particular URL and the frequency of keywords on weighted pages.

According to Kruger, the strength of any keyword or resource on the user's page determines its chances of survival when the page is merged with that of another Web user.

"The software can be set to a default where information searched on the Web only once will not be listed. Only if GenePage sees a resource being revisited will it be listed in their resource pages. If other users in a GenePage group also access the resource, then that URL will rise up the page in relation to its importance.

According to Kruger, GenePage is aimed at helping people work more effectively in a distributed organization. As the network of GenePages develops the data they contain converges on the resources and information best suited to the workgroup they belong to.

This, he explained, makes GenePage ideally suited for collaborative working and a range of intranet based applications such as medical and scientific research where grading the potential sources of information helps users make informed decisions.

Steinkrug's Web page is at http://www.steinkrug.com.