Relevant links
Relevance isn’t the thing which lives in HTML-document by itself. Relevance is a coefficient of the conformity with HTML-document query. Relevance, counted by search engines, is a very subjective thing because of an imperfection of algorithms and limits of search engines.
Every search engine defines the relevance of HTML-document to the user’s query according to the searching concept which lies in it. Although the conceptions are different, search engines search in the same way because search algorithms are built on the common principles. Main differences of search engines are not in algorithms of relevance definition, in their realization.
In catalogues relevance is evaluated by people (moderators), their duty is to make recourse sorting according to sections and cutting off spam. When user makes query, search engine of a catalogue counts relevance, as the search engine of a search system does, but considering the evaluation of moderators.
Many different factors influence the search system evaluation, beginning from the domain name and ending with the quality of connection channels. Further the factors, which are defied evaluation and control, are enumerated, which influence the relevance of HTML-documents:
· – Domain name.
· – Tag <Title>.
· – META tags, <Keywords>, <Description>.
· – META tags <Robots>, file robots.txt (or its absence).
· – META tags <Refresh>, with value close to zero.
· – META tags <Expires>, if past date is given.
· – META tags <Document-state>, defines indexation regime.
· – HTML-code size, which is before text.
· – Mistakes in code.
· – Mistakes in text.
· – Text size.
· – Text quality (style, content, claiming).
· – The amount of key wards in HTML-document.
· – Moving away of key wards from the text beginning.
· – Grouping of key wards.
· – Exact accordance to a key phrase.
· – Key wards separation. Tags <B>, < H1 > – < H6 >, <STRONG>.
· – Tags <AREA>, <IMG>.
· – Tag <A>.
· – Tag <FRAME>.
· – Tag <SCRIPT>.
· – Tag <!-commentary tag–>.
· – Tags <STYLE>, <BODY>, <FONT>, <TABLE>.
· – The size of images placed in documents.
· – The amount of documents on server.
· – The amount of “quality” documents on server.
· – The amount of “garbage” on server.
· – The organization of links within server (depth, coverage, amount).
· – Inner links to server pages (amount).
· – Rating and accordance to server theme, to which are outer links come.
· – Popularity of server (the amount of adverts).
· – The time of server life (page), the more the better is.
· – The organization of links within server.
· – The quality of a virtual server, on which recourse is placed (traffic limits, reliability).
Very important but uncontrolled factors:
· – The level of competition in chosen thematic.
· – The quality of recourses in chosen thematic.
· – Claiming of chosen thematic.
· – The amount of spam, attacking search system in chosen thematic.
