When customers search for products in online stores, their previous behavior and interests are mostly ignored. The relevance of search results can be significantly improved if, in addition to the query, the intention of the users is taken into account. mgm-experts Liliya Avdiyenko and Christian Winkler demonstrate in the article “Echtzeitkontext für bessere Suchergebnisse auf Websites” (Real-time context for better search results on websites) – published by “heise developer” – how this works in practice. The authors describe a method which predicts the search intention of users. It is based on historical log data of the website and the real-time search context, which contains, among other things, information about the current session. With the aid of an algorithm for machine learning, individualized search results can be generated that take into account the user’s intentions and the context of the respective session.