Taggis - Attribute extraction Engine for attribute search

Taggis® ... is a product/offer characteristic extraction engine especially conceived for the e-commerce field. It opens a new era in searching for products on the Internet. Taggis  is based on Sight’Up’s artificial intelligence technology. It is self-learning and works on the principle of "Machine-Learning".

Innovative Technology : Taggis makes it possible to determine the principal characteristics of a product or offer. The corresponding  search results are thus extremely precise. Taggis also makes it  possible to propose similar products at a simple clic from a list of attributes (for example: similar products in another material).

Let us take an extract from a product :

…this charming red silk dress supplied with a black belt ….

In the above text traditional technologies will present this product as a result of the following research : "black dress" whereas Taggis won't, estimating it does not correspond to the request of the visitor.

Very little human intervention : the persons charged with setting up and maintaining Taggis do not need any particular IT skills. They just need to know their area of business well.

• Just 10 learning offers are enough to make it possible to manage more than 20 000 offers.
• The characteristics of attributes are managed by glossaries. Sight’Up also proposes the Dis engine for helping in the construction of these glossaries.

Unlike other tagging systems, Taggis requires neither rules of management, nor dictionary ; thus allowing its adaptation to any type of corpus or language. The integration of new examples is thus carried out automatically, without any human intervention.

Taggis manages daily more than 100 million product offers in more than 10 languages for Kelkoo (European leader of shopping engines - shopbot and subsidiary of Yahoo)

Taggis has been entirely developed in C in order to provide a large processing capacity and universal portability. It is supplied with its integration APIs and works in XML format and Unicode.


More details on Taggis