The project concerns the employment of real-time analytics in data streams emitted from a mobile phone tracking system in a large supermarket in Greece.
Research & Development
In the context of this project the following were performed:
- analysis of the data already collected from an developed tracking system, both concerning its structure and by applying clustering to discover possible usage scenarios,
- a survey of the layout of the store and the locations of the tracking mechanisms, in order to create a correlation matrix of the detections and clarify the results of the analysis,
- specification, design and development of a real-time machine learning system capable of recognising the fingerprint of the movement of customers within the supermarket and
classifying it in various classes.
Through this project:
- Degrees of similarity between users can be calculated, allowing grouping the users in categories of shoppers who, when visiting the store, take a similar route and stay there
for a similar time.
- Shopping “missions” can be identified (i.e. fast moving customers that pick up some targeted products).
- Focus points of customers were able to be detected.
Please contact us for more information on this project.