Predicting commercial activeness over urban big data
Title
Predicting commercial activeness over urban big data
Subject
Forecasting
Big data
Planning
Urban planning
Population statistics
Ecosystems
Zoning
Taxicabs
Crowdsourcing
Description
This study aims at revealing how commercial hotness of urban commercial districts (UCDs) is shaped by social contexts of surrounding areas so as to render predictive business planning. We define social contexts for a given region as the number of visitors, the region functions, the population and buying power of local residents, the average price of services, and the rating scores of customers, which are computed from heterogeneous data including taxi GPS trajectories, point of interests, geographical data, and user-generated comments. Then, we apply sparse representation to discover the impactor factor of each variable of the social contexts in terms of predicting commercial activeness of UCDs under a linear predictive model. The experiments show that a linear correlation between social contexts and commercial activeness exists for Beijing and Shanghai based on an average prediction accuracy of 77.69% but the impact factors of social contexts vary from city to city, where the key factors are rich life services, diversity of restaurants, good shopping experience, large number of local residents with relatively high purchasing power, and convenient transportation. This study reveals the underlying mechanism of urban business ecosystems, and promise social context-aware business planning over heterogeneous urban big data. 2017 Association for Computing Machinery.
3
1
Creator
Yang, Su
Wang, Minjie
Wang, Wenshan
Sun, Yi
Gao, Jun
Zhang, Weishan
Zhang, Jiulong
Publisher
Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies
Date
2017
Type
journalArticle
Identifier
24749567
10.1145/3130983
URL
http://dx.doi.org/10.1145/3130983
Citation
Yang, Su et al., “Predicting commercial activeness over urban big data,” Lamar University Midstream Center Research, accessed May 18, 2024, https://lumc.omeka.net/items/show/28498.