Abstract
To stay competitive, plenty of data mining techniques have been introduced to help stores better understand consumers' behaviors. However, these studies are generally confined within the customer transaction data. Actually, another kind of 'deep shopping data', e.g. which and why goods receiving much attention are not purchased, offers much more valuable information to boost the product design. Unfortunately, these data are totally ignored in legacy systems. This paper introduces an innovative system, called TagBooth, to detect commodities' motion and further discover customers' behaviors, using COTS RFID devices. We first exploit the motion of tagged commodities by leveraging physical-layer information, like phase and RSS, and then design a comprehensive solution to recognize customers' actions. The system has been tested extensively in the lab environment and used for half a year in real retail store. As a result, TagBooth generally performs well to acquire deep shopping data with high accuracy.
Original language | English |
---|---|
Title of host publication | 2015 IEEE Conference on Computer Communications, IEEE INFOCOM 2015 |
Publisher | IEEE |
Pages | 1670-1678 |
Number of pages | 9 |
Volume | 26 |
ISBN (Electronic) | 9781479983810 |
DOIs | |
Publication status | Published - 21 Aug 2015 |
Externally published | Yes |
Event | 34th IEEE Annual Conference on Computer Communications and Networks, IEEE INFOCOM 2015 - Hong Kong, Hong Kong Duration: 26 Apr 2015 → 1 May 2015 |
Conference
Conference | 34th IEEE Annual Conference on Computer Communications and Networks, IEEE INFOCOM 2015 |
---|---|
Country/Territory | Hong Kong |
City | Hong Kong |
Period | 26/04/15 → 1/05/15 |
Keywords
- Action Recognition
- Deep Shopping Data
- Motion Detection
- RFID
- TagBooth
ASJC Scopus subject areas
- General Computer Science
- Electrical and Electronic Engineering