Home // eKNOW 2023, The Fifteenth International Conference on Information, Process, and Knowledge Management // View article

Mining User Behavior: Inference of Time-boxed Usage Patterns from Household Generated Data

Authors:
Ramona Tolas
Raluca Portase
Mihaela Dinsoreanu
Rodica Potolea

Keywords: knowledge inference; clustering; event-based signal processing; pattern mining; household-generated data; usage mining

Abstract:
The growth of technology and the reduced cost of data storage are enablers for producing and storing a large amount of data. Smart household devices are a category of data producers due to the monitoring sensors equipping the device. These sensors monitor the device's state and user interaction with the device. Besides the initial reason for planting the monitoring equipment, further valuable information can be extracted from this data, such as user behavior. Mining usage patterns can further be used in forecasting user presence, data-driven decisions, or service personalization. A processing pipeline for mining usage patterns is proposed in this paper. The problem is theoretically formulated and a method of mining usage patterns is proposed. The method is developed and tested on synthetic data and interesting insights are extracted from real data by deploying the pipeline in a data lake containing real interactions of users with smart home appliances. Similarities between real usages of household appliances are found as a result of this step and several categories of users are defined based on them.

Pages: 50 to 58

Copyright: Copyright (c) IARIA, 2023

Publication date: April 24, 2023

Published in: conference

ISSN: 2308-4375

ISBN: 978-1-68558-082-7

Location: Venice, Italy

Dates: from April 24, 2023 to April 28, 2023