Theoretical Aspects

From RankAnalyst Lab
Jump to navigation Jump to search

As long as we assume that the search results of the search engines are not compiled purely randomly, an exact examination of the search results is of particular importance. The results reveal a lot about the algorithm and thus the "understanding" that the search engine has of the website being searched. Since Google is an intelligent mediator between supply (websites) and demand (searchers), it is also possible to derive information about the searcher, the target group of a website.

Term Vector Model

When we talk about "understanding," we need to look at the basic technology that Google uses to classify content in terms of its meaning. For this the so called Term Vector Model (or "Vector Space Model"), is employed which allows Google to locate identical or similar contexts (search expressions) in a multidimensional space. If the expressions are geometrically close to each other, their content is also more closely related. This technology enables Google to accurately classify the search query, but also to better categorize website content or pre-classify it.

User signals/Interaction

The question of whether and to what extent users influence Google's ranking through their behavior is a controversial one. In particular Google itself does not express itself unequivocally here, partly contradictory. In daily monitoring, we repeatedly find clear indications as to how Google captures user signals and incorporates them into the ranking.

Since the effect varies from case to case, interaction data should be considered regularly in each case. Suitable analysis options are available in RankAnalyst Lab.

More "understanding" through AI and RankBrain

Google also relies in ( key ) parts on the use of neural networks (AI) in order to get even closer to people in terms of terminological understanding. The so-called RankBrain algorithm should enable a targeted, dynamic weighting of different ranking factors and thus derive an even better context.

Only if we know what Google thinks about them can we take appropriate measures to positively influence the profitability of a website. And since the effects of the Google AI, but also all other mechanisms, are reflected in the search results, we also have to search for their patterns there. RankAnalyst Lab offers the appropriate tools for this purpose.

Playing Field/Reserve Bench

Spielfeld/Reservebank-Axiom