Unique Concept of Human and Artificial Operating Synergy – UnChaos. [see: chaos theory]

Home

Products & Services

Technology

Publications

About



...Unlike classical hypertext, the Web is constantly changing... the change on the Web gives it an extra dimension – time – that can be exploited by accessing old data stored by search engines...”
-- Ellen Spertus, Mining Structural Information on the Web. Constant change, 1997.

There are numerous efforts to describe the dynamic nature of the web, but most of them do not sufficiently handle the web's sociological aspect, and as a result they could not exploit the web dynamic as a measure of a web page relevancy. Those insufficiencies stem from inability of understanding of meant ideas by producing the extrinsic text filtering only, in other words dealing with textual elements and not with their meanings.
Considering a web dynamic, first of all, we associate the web with a semi-organized society and its history, or at its different periods of being. In this light, a tracking history of web changes does not only mean a tracking of modifications of web pages or web sites and saving the recent ones, it is like tracking a human history with weaving different episodes into a whole scenario using various types of identifications, or names. And like a human history, where some names play the major role in highlighting of various events or even epochs, and some of them are just secondary facts that were caused by those days events, at the web can be seen the same picture, where some names are the representative for a topic, or build its popularity, and some of them became popular because of being considered at a popular topic.
Based on the above we can define the web dynamic as verifying and tracking a popularity of web topics and names, this means verifying names, tracking number of their appearances (we call them as Name Soft Links), and relating them to relevant topics.
However, there are some obstacles to the exploitation of the web dynamic as a measure of relevancy. The major one is enormous difference between “old” names and “new” ones. For example a new chip maker company, which increased appearance of their product name on the web 10 times at the last decade, has relatively small amount of this name appearances yet, while an old chip maker company has relatively large number of their chip name appearances even after decreasing. In order to improve this situation we define a name progress tendency, which means taking into account the history of changes of name appearances, with their increasing or decreasing tendencies for a period of time, for estimating of web pages relevancies.
Another aspect of web dynamic is a popularity of its topics. A topic become as more popular as more appearances of related to it names, on different web sites, are met on the web. The same way, as was defined a name progress tendency, is defined a web topic progress tendency. In other words for a given period of time it is verified an increasing or decreasing tendency of popularity of a topic, it is done by estimating of a delta (difference) between appearances of names related to a topic for a period of time. For example, in the last decade we see an increasing tendency in producing cellular phones, many companies decided to enter this market, and as a result on the web appear a lot of new product names which are related to cellular phones production topic. In other words a cellular phones production topic on the web has a positive progress tendency at the last decade. This measure can be used for ranking topics related to a user search engine request.



© UnChaos Ltd.