http://www.scipool.com/doc/doc200410284501.html

Click here to read the complete document

Introduction

Let’s see the Realm where these operators will perform their mission. One world wide realm will be the Web space. In the Web space billion of documents are hosted that are indexed and referred in some special Websites known as "Search Engines". The Web space is far from being considered a Library because documents are hosted in sites at will, without any type of permission and control. The Web is absolutely free. Of course Websites are more or less organized, more or less specialized and with different levels of popularity and traffic. Knowledge is dispersing all over Web, no "mapped" at all. Concerning a specific discipline (Major Subject of the Human Knowledge), for instance medicine, we may find via Search Engines more than 40 million of documents. Why so much?. Because any of those documents referred have the "word" medicine cited one or more times within their content. From time to time appears in the Web Community some knowledge organizers known as "Virtual Libraries" that try to make a meaningful Cognitive Offer of the main authorities of a given discipline. Most of them are voluntary efforts that accept the not always bona fide Websites owner’s promotion of their respective contents. These Virtual Libraries generally offer only a few thousands of references. To provide more and better information to robots that continuously crawl the Web space gathering references for their respective Search Engines some special command were added to HTML like Meta Tags, where Web page owners are allowed to include some crucial "keywords" supposedly present in the document as an aid to guide users’ preferences and needs.

Unfortunately these Meta Tags are frequently wrongly used to mislead people. As a better approach some Search Engines proceed then to browse by themselves document by documents parsing them as analytically as possible, from their beginning to their end. In this way the indexing task is more objective. They do even more in order to provide people better information about sites: their "popularity", a measure of the relative importance of them in the Web space. As a huge democratic Community they alleged that "voting democratically" among them, would be the best possible way to measure their relative authoritativeness. A vote is a mention of a given site. It is reasonable that sites that are most referenced by the rest of the Community could be considered the most authoritative even though with a popularity metric.

Click here to read the complete document