Big Data for Urban Analysis

  • General info (Source: Osiris)
  • Quartile: 3-4
    Time Slot: E
    Course Type: Specialization Elective
    Code: 7ZW1M0
    Responsible Lecturer: A.D.A.M. Kemperman
    ECTS: 10
    Exams: No
    Required courses:


    Course description:

    To find good solutions one need to have a good understanding of the problem. This holds true also for the problems urban planners are facing in areas such as mobility (congestion and accessibility), health (air pollution, passive life styles), energy (smart grids and transformation to renewable sources of energy), ageing (social exclusion, social satisfaction), and tourism (crowding). In this project you consider a planning problem of your choice and apply a suitable approach to better understand the problem and evaluate scenarios.
    The approach includes Information from a big database such as GPS data, Twitter data or one of the large national surveys, such as OVIN. These databases provide rich information on micro-level of individuals. In this approach an existing database, or combination of databases is analysed to achieve a better understanding of behaviour of individuals with regard to the planning problem considered. During the project the following steps will be carried out: formulation of a research question; literature research; specification of a conceptual model; identification of relevant variables; preparation of the data; performing the analysis and interpreting the results. The analysis technique and database used will be chosen depending on the research question. The emphasis is on advanced techniques from the field of either regression modelling (e.g., path analysis) or data mining (e.g., Bayesian network learning).

  • Other courses recommended by students
  • Useful preliminary courses:

    Urban Research Methods (Strongly recommended)

    Smart Urban Environments

    Useful follow-up courses:

    The project does not allow for expansion, but helps to gain knowledge for SPSS or data-management related courses

  • Survey outcomes (0: low, 5: high)
  • Time input: You need to be logged in to view this content
    Difficulty: You need to be logged in to view this content
    Fun to do: You need to be logged in to view this content
    Interesting: You need to be logged in to view this content
    Application to practice: You need to be logged in to view this content
    Gained knowledge: You need to be logged in to view this content
    What did people like (in year 2020-2021): You need to be logged in to view this content
    What can be improved (in year 2020-2021): You need to be logged in to view this content

  • Additional information
  • Applied skills / methods:

    Bayesian Belief Networks (BBN)

    Project course: Yes
    Chosen topics:
    • Influence of travel behaviour on air quality
      • Using a Bayesian Belief Network,  the influence of social-demographical and environmental characteristics on travel behavior and then the influence of this is on the air quality are researched, while also including traffic data.
    • Urban data management
    • What influences the mode choice when people conduct a shopping trip
    Recommended topics:

    You could pick your own topic, there was no predefined list. So choose whatever you would like to do

    • But do keep in mind you do not choose to similar topics to other students, so that your presentation and outcomes are more interesting
    • Especially traffic was found to be a common subject

    Try researching with social media data. You might not get a second chance during the master to try it out elsewhere


  • Metadata
  • Data source: Own survey
    Applied method: Questionnaire
    Response rate: 18%
    Sample size: 5
    Academic year: 2020-2021

  • Disclaimer: The following data has been collected by SERVICE among students that followed this course in academic year 2020/2021. Based on this feedback or other causes, it is possible that the course will have a different set up in the future. Keep this in mind when you use these data for selecting your courses. Additionally, due to the low absolute sample size (5), the opinions of the students may not correctly represent the opinion of the full class that attended the course