Recognition of Urban Functional Areas Based on Call Detail Records and POIs Data

As the emergence of big data in recent years, there have been lots of transformations going on the study of big data mining, which brings the rapid development of urban computing. Meanwhile, the integration of planning and management in urban area is raising higher level of requirements for acquiring dynamic Urban Functional Areas (UFA), which encourages the researches on identification of dynamic UFAs. This paper presented the concept of Communication Activity Intensity (CAI) and deemed it is much more meaningful than call volumes on the research of UFAs recognition. Besides, diverse Basic Research Units (BRU) were discussed about their impacts on the accuracy of recognition before modeling. Moreover, the k-means clustering method for dynamic Call Detail Record (CDR) data and Kernel Density Estimation (KDE) model for static Point of Interest (POI) were established at the Traffic Analyses Zone (TAZ) level. At the same time, the case study on the region within the 3rd Ring Rd. of Beijing was conducted and the results of identification were checked by qualitative and quantitative verifications. Finally, the highest identification index 96.00% on the function of park & scenery area was obtained, and the average value is 77.05%. In addition, residential area is 82.50% and office area is 75.76%. In conclusion, the accuracy improved comparing with previous researches, which verifying the reliability of the method.

  • Supplemental Notes:
    • This paper was sponsored by TRB committee ABJ70 Standing Committee on Artificial Intelligence and Advanced Computing Applications.
  • Corporate Authors:

    Transportation Research Board

    ,    
  • Authors:
    • Yuan, Guang
    • Chen, YanYan
    • Sun, Lishan
    • Lai, Jianhui
    • Liu, Zhuo
  • Conference:
  • Date: 2019

Language

  • English

Media Info

  • Media Type: Digital/other
  • Features: References; Tables;
  • Pagination: 5p

Subject/Index Terms

Filing Info

  • Accession Number: 01697457
  • Record Type: Publication
  • Report/Paper Numbers: 19-03732
  • Files: TRIS, TRB, ATRI
  • Created Date: Dec 7 2018 9:28AM