To search, Click below search items. |
| All Published Papers Search Service |
|
|
|
Title |
Improving Shopping Experience of Customers using Trajectory Data Mining |
|
Author |
Tanuja. V, Govindarajulu. P |
| Citation |
Vol. 17 No. 2 pp. 170-180 |
|
Abstract |
The up-to-date tracking tools, such as RFID (Radio Frequency Identification), Wi-Fi sensing brought a tremendous breakthrough in business analytics. These current trends can be used to record the paths taken by shoppers in a supermarket which form a trajectory data. By mining these trajectories it is possible to uncover shopper’s behavior including how they move among shelves of different product types, the time they spend on specific shelves, and other information that can be used to make a product arrangement that improves customers’ shopping experience and hence attract customers. Trajectories of shoppers in a grocery store as recorded from RFID tags or with the records of attendant located on their shopping ride forms the basis for the analysis. By using a clustering algorithm, we can profile shopping paths by neighborhoods visited and then by time spent to uncover the common paths of different categories of shoppers such as those who are looking to grab a few important items and leave. The clustered shoppers’ trajectories of items uncover the customer behavior that can be used to sketch groups of shoppers based on their space use and assess the impact of the spatial configuration of a store layout on shoppers’ behavior. In this paper, we propose a novel method of shopping super market items of trajectory clustering algorithm which groups similar transactions of super market item sets (trajectory) for shop which elucidates customer shopping behavior. A new similarity measure between transaction trajectories of super market items to group market basket item sequences of trajectories is developed. Our proposed method can find the dominant super market shopping path sequences of item sets that are capable of identifying the hotspots where most of the customers’ visits are made and the least visited paths as well. This can understand shopper behavior in a store. We used the real dataset of a supermarket in Nellore to apply the proposed methodology, as case study, to demonstrate the advantages and usefulness of the method. |
|
Keywords |
Data mining, Market Basket Analysis, Clustering, LCSS clustering, Sequence of items, purchasing patterns |
|
URL |
