A Smart Shopping Assistant utilising Adaptive Plan Recognition

In this paper we describe an adaptive shopping assistant system utilising plan recognition. Radio Frequency Identification (RFID) sensory is used to observe a shopper’s actions, from which the plan recogniser tries to infer the goals of the user. Using this information, an automated assistant provides help tailored to the shopper’s concrete needs. We discuss why it is crucial to make the plan recognition process itself user adaptive and present ideas how to realise this through modification of existing plan recognition approaches.