The knowledge graph is a phrase that is being used more frequently to indicate a spatial relationship between things. This is where ‘search’ for information is heading.
It is an interesting concept, linking together two concepts in information theory.
Firstly it considers what we mean by words such as ‘bath’. Is this a town in England, an item of furniture you purchase at a plumber, or are you looking for historic Roman Baths or a Spa? The answer is there are many possibilities and by showing these to the searcher, the required path can be followed to the desired object, not the desired keyword. This will soon have a major impact on Google search – it already does in some other language search engines.
But it also highlights that it is the object – the ‘thing’ – that is important – again this is important when dealing with many language or dialects. Many words might be used for the same object.
Secondly it is looking at the structure and relationships between ‘things’, enabling the searcher to find other similar objects. This impacts how search results can be shown – and many search engines are changing from ones set of ‘results’ to an array of different results. This is most developed in some Asian search engines, but also the Russian Yandex. Google recently started a similar change in search results, with many more changes expected shortly. The key thing here is they are trying to structure the results.
An earlier ExtraDigital article by Rachel Cornish looked at how search engines are going ‘back to basics’ or back to their roots and utilising the real information within search and not just keywords.
So the knowledge graph is an important concept for all those involved in search marketing.
