Modelling Behavior for the Optimization of Mobile Interactions and Supply--Demand Alignment for Sharing Economy Platforms.

Saved in:

Bibliographic Details
Title:	Modelling Behavior for the Optimization of Mobile Interactions and Supply--Demand Alignment for Sharing Economy Platforms.
Authors:	Ma, Zhaohe¹ hedy_mazhaohe@163.com
Source:	International Journal of Interactive Mobile Technologies. 2026, Vol. 20 Issue 14, p117-131. 15p.
Subjects:	Human behavior models, Supply & demand, Mobile operating systems, Edge computing, Ridesharing services, Selectivity (Psychology)
Abstract:	Shared mobility platforms rely on mobile terminals to support large-scale service scheduling. However, the limited computational resources of mobile devices and the reliance of existing approaches on trajectory data alone have resulted in insufficient demand perception accuracy and elevated supply--demand matching latency, primarily because dynamic user interaction behaviors have been overlooked. To address these limitations, a progressive and integrated framework encompassing mobile sensing, behavior modeling, and edge-assisted matching was developed to enable fine-grained user behavior modeling and realtime supply--demand optimization in shared mobility environments. Mobile interaction entropy and conditional interaction entropy were introduced to quantify the degree of operational disorder and the temporal regularity of user behaviors, respectively. A lightweight dual-stream attention network was designed to fuse trajectory-based spatial features with interaction-behavior features, thereby facilitating accurate short-term travel demand prediction. A weighted loss function was incorporated to strengthen the fitting performance of difficult-to-predict scenarios. Within an edge computing architecture, dynamic contextual matching criteria were established through online metric learning with positive-definite constraints, while locality-sensitive hashing was employed to accelerate candidate retrieval and achieve millisecond-level supply--demand matching. Furthermore, a closed-loop edge--cloud collaborative updating mechanism was developed to balance local environmental adaptability with global model generalization. [ABSTRACT FROM AUTHOR]
	Copyright of International Journal of Interactive Mobile Technologies is the property of International Journal of Interactive Mobile Technologies and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database:	Engineering Source

Description
Abstract:	Shared mobility platforms rely on mobile terminals to support large-scale service scheduling. However, the limited computational resources of mobile devices and the reliance of existing approaches on trajectory data alone have resulted in insufficient demand perception accuracy and elevated supply--demand matching latency, primarily because dynamic user interaction behaviors have been overlooked. To address these limitations, a progressive and integrated framework encompassing mobile sensing, behavior modeling, and edge-assisted matching was developed to enable fine-grained user behavior modeling and realtime supply--demand optimization in shared mobility environments. Mobile interaction entropy and conditional interaction entropy were introduced to quantify the degree of operational disorder and the temporal regularity of user behaviors, respectively. A lightweight dual-stream attention network was designed to fuse trajectory-based spatial features with interaction-behavior features, thereby facilitating accurate short-term travel demand prediction. A weighted loss function was incorporated to strengthen the fitting performance of difficult-to-predict scenarios. Within an edge computing architecture, dynamic contextual matching criteria were established through online metric learning with positive-definite constraints, while locality-sensitive hashing was employed to accelerate candidate retrieval and achieve millisecond-level supply--demand matching. Furthermore, a closed-loop edge--cloud collaborative updating mechanism was developed to balance local environmental adaptability with global model generalization. [ABSTRACT FROM AUTHOR]
ISSN:	18657923
DOI:	10.3991/ijim.v20i14.62555