![Page 1: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/1.jpg)
Urban Computing
Dr. Mitra Baratchi
Leiden Institute of Advanced Computer Science - Leiden University
14 February, 2020
![Page 2: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/2.jpg)
Agenda for this session
I Part 1: Practical mattersI Part 2: Introduction to the course
I What is Urban ComputingI ApplicationsI Data sources
I Part 3: Hands-on Lab
![Page 3: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/3.jpg)
Part 1: Practical matters
![Page 4: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/4.jpg)
Teaching assistants
![Page 5: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/5.jpg)
Courseware
I Courseware access:https://urbancomputing2020.github.io/
I Other matters (Announcements, assignment hand-in,discussion forum): Blackboard
![Page 6: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/6.jpg)
Communication
Before sending emails:
I Can you ask the question during the class?
I Can you use blackboard forum?
![Page 7: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/7.jpg)
Course schedule
Work group session (11:15-13) Lecture (14:15-16) 16:15-18 Deadlines 7 Feb 2020 - No lecture (Dies) - - 1 14 Feb 2020 - Introduction - 2 21 Feb 2020 - Time series data processing,
Assignment 1 - Deadline selecting a paper
3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal data processing - Deadline Assignment 1(9 March, 23:59) 5 13 Mar 2020 - Machine learning for urban
computing, Assignment 2 -
- 20 Mar 2020 Assignment 2 - - 6 27 Mar 2020 Assignment 2 Machine learning for urban
computing 2 - Deadline Assignment 2 (30 Match, 23:59)
- 03 Apr 2020 - No lecture (data science exam) - - 10 Apr 2020 - No lecture (Good Friday) - Deadline proposal (10 April, 23:59) 7 17 Apr 2020 - Data visualization for urban
computing -
- 24 Apr 2020 Meeting with teams (with appointment)
- -
8 01 May 2020 Meeting with teams (with appointment)
- -
9 08 May 2020 Presentation Presentation - 10 15 May 2020 presentation Presentation Presentation Peer review 11 12 June 2020 - - - Deadline project (15 June 23:59) 12 10 July 2020 - - - Resit project, assignments with penalty
(10 July 23:59)
Individual meeting with teams will be scheduled later
![Page 8: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/8.jpg)
Organization of class
Course structure
I Lectures
I Practical sessions for assignments (According to the schedule )
I Practical lab material (self-study material)
I Feedback on projects
I Presentation by students
![Page 9: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/9.jpg)
Grading
I Active participation in class and discussions (10%)
I Assignments (25 %)
I Presentation (15 %)I Project (50 %)
I Novelty of the ideaI Maturity of experiments (considering availability of data
sources, or source codes)I ResultsI Presentation and documentation
![Page 10: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/10.jpg)
Course rules
I To pass the course both the combined grade and theassignments grade should be over 5.5
I Deadlines are fixed → no late submission
I Late submission is considered as resit
I Resit option for assignments and project has maximum gradeof 7
I No resit option for presentations
I Submitting project proposal is compulsory (no submissionimplies deduction of 2 points from the project)
![Page 11: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/11.jpg)
Project
I Select a reference paper as a starting pointI Bid on the paper bidding list available (select top 5 papers)I Alternatively you can bid for papers cited in the Urban
computing survey paper [ZCWY14], or previously acceptedpapers to the annual Urban Computing workshop any year
I Register your bid (in the form we provide on blackboard)
I Write a brief proposal including (problem statement, researchquestion, methodology, evaluation approach, data sources)
I Proceed with the project
I Write a report (6-8 pages (ACM-proceedings Latex template))[download]
![Page 12: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/12.jpg)
Part 2: Introduction Urban Computing
![Page 13: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/13.jpg)
What does Urban Computing Mean?
1
Figure: Urban Computing
1source: http://uctutorial.chinacloudsites.cn
![Page 14: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/14.jpg)
The familiar stranger... [PG04]
![Page 15: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/15.jpg)
A bit of history ....
When was it mentioned first?
”The term urban computing was first introduced by Eric Paulos atthe 2004 UbiComp (Ubiquitous and pervasive computing)conference”
![Page 16: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/16.jpg)
My own story as a member of Pervasive Systems
Pervasive and ubiquitous computing (or ”ubicomp”) is a conceptin software engineering and computer science where computing ismade to appear anytime and everywhere. In contrast to desktopcomputing, ubiquitous computing can occur using any device, inany location, and in any format
2
2source:https://www.parc.com/wp-content/uploads/2010/03/ubicompvenn.jpg
![Page 17: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/17.jpg)
Ubiquitous Computing research with the focus on mobilitydata
![Page 18: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/18.jpg)
Back to Urban Computing
Urban Computing is the process of acquisition, integration, andanalysis of big and heterogeneous data generated by diversesources in urban spaces, such as sensors, devices, vehicles,buildings, and humans, to tackle the major issues that cities face(e.g., air pollution, increased energy consumption, and trafficcongestion). Urban computing connects unobtrusive andubiquitous sensing technologies, advanced data management andanalytic models, and novel visualization methods to createwin-win-win solutions that improve urban environment, human lifequality, and city operation systems. [ZCWY14]
![Page 19: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/19.jpg)
Mention some urban computing applications ....
![Page 20: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/20.jpg)
Applications
3
Figure: Traffic management
3source: https://www.autoevolution.com/news/the-longest-traffic-jam-in-history-12-days-62-mile-long-47237.html
![Page 21: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/21.jpg)
Applications
Figure: Event management
![Page 22: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/22.jpg)
Applications
4
Figure: Autonomous driving
4source: http://senseable.mit.edu/light-traffic/
![Page 23: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/23.jpg)
Figure: Urban planning
5
5source:[vSTGB]
![Page 24: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/24.jpg)
Rewilding
6
Figure: Rewilding
6source:https://www.ark.eu/gebieden/buitenland/rewilding-europe
![Page 25: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/25.jpg)
Example of rewilding in the Netherlands
Figure: Rewilding
![Page 26: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/26.jpg)
Main categories of applications in Urban Computing
7
7image source:[ZCWY14]
![Page 27: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/27.jpg)
What do we learn in this course?
I Things a computer scientist should know when using data tosolve urban problems
![Page 28: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/28.jpg)
What do we learn in this course?
I Things a computer scientist should know when using data tosolve urban problems
![Page 29: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/29.jpg)
What we, unfortunately, won’t learn in this course:
I Urban planning
I Urban policy making
I Urban ethics
I ...
![Page 30: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/30.jpg)
Topics
I Data sources for Urban Computing research
I Processing time-series data
I Processing spatial data
I Processing spatio-temporal and trajectory data
I Machine learning algorithms for Urban Computing research
I Deep learning for Urban Computing research
I Visualization techniques for Urban Computing research
![Page 31: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/31.jpg)
Why Urban Computing as a new field?
Thinking about urban problems is not new, people have collecteddata to solve these problems since a long time ago....
![Page 32: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/32.jpg)
Data used for solving urban problems
I Old data sources
I Modern data sources
![Page 33: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/33.jpg)
Old data sources
I Calling people on phones for calculating origin destinationmatrices (traffic engineering)
I Questionnaires
I Census
I Observations by social scientists
![Page 34: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/34.jpg)
Modern data sources [AB14, AB18]
Modern data sources are categorized into the following threecategories based on the origin of the data:
I Bottom up: Citizens
I Intermediate: Digital companies
I Top down: Government
![Page 35: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/35.jpg)
Buttom-up: citizens as sensors
I Data collected through sensing phones (in some manner)
I Data generated as a result of using Apps
I Participatory sensing: ( communities (or other groups ofpeople) contributing sensory information to form a body ofknowledge)
![Page 36: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/36.jpg)
Buttom-up: citizens as sensors
I Data collected through sensing phones (in some manner)
I Data generated as a result of using Apps
I Participatory sensing: ( communities (or other groups ofpeople) contributing sensory information to form a body ofknowledge)
![Page 37: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/37.jpg)
Buttom-up: citizens as sensors
I Data collected through sensing phones (in some manner)
I Data generated as a result of using Apps
I Participatory sensing
: ( communities (or other groups ofpeople) contributing sensory information to form a body ofknowledge)
![Page 38: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/38.jpg)
Buttom-up: citizens as sensors
I Data collected through sensing phones (in some manner)
I Data generated as a result of using Apps
I Participatory sensing: ( communities (or other groups ofpeople) contributing sensory information to form a body ofknowledge)
![Page 39: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/39.jpg)
Ways to collect data by localizing phones
8
Figure: Sensing movement using cellular networks
8source: http://unbonmotgroundswell.blogspot.com/2013/07/hybrid-location-technologies-gps.html
![Page 40: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/40.jpg)
Wifi sensing
Figure: Sensing movement using Wifi networks [PCB+17]
![Page 41: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/41.jpg)
Wifi sensing, and privacy
9
Figure: Wifi sensing
9source: https://obj.ca/article/techopia-ottawas-edgewater-wireless-unveils-wi-fi-location-tracking-tech
![Page 42: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/42.jpg)
Figure: Assen sensor setup
![Page 43: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/43.jpg)
Intermediate: digital companies
I Free services provided by companies through Internet
I Data generated as a result of the side activity of a digitalcompany
I Companies that aggregate data from local brokers (Funda)
![Page 44: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/44.jpg)
Funda.nl
Figure: Funda
![Page 45: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/45.jpg)
Foursquare
Figure: Foursquare
![Page 46: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/46.jpg)
Top-down: government (Open Data)
Open data is the idea that some data should be freely available toeveryone to use and republish as they wish, without restrictionsfrom copyright, patents or other mechanisms of controlGovernment institutions release (part of) their internal data inopen format.
![Page 47: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/47.jpg)
Dutch open data portal
Figure: Dutch open data portal.
![Page 48: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/48.jpg)
A dataset search engine (to be tested)
Figure: Google dataset search
![Page 49: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/49.jpg)
Old and modern sources (comparison)
I Where to start from? Collecting data for a specific researchquestion or finding a research question based on data?
I Data collection costs
I Data granularity
I Data quality (noise, error, etc)
I Data sparsity (duration period, missing data)
I Inference quality (one person holding two phones is countedtwice,... )
![Page 50: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/50.jpg)
Machine learning/data mining versus statistics
Current approaches to spatio-temporal data handling:
I Statisticians approach
I Machine learning approach
![Page 51: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/51.jpg)
Statisticians approach
Statistics is a branch of mathematics dealing with data collection,organization, analysis, interpretation and presentation [Wikipedia]
I Hypothesis testing
I T-test
I Permutation test
I ...
![Page 52: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/52.jpg)
Machine learning/data mining approach
Machine learning (ML) is the scientific study of algorithms andstatistical models that computer systems use to effectively performa specific task without using explicit instructions, relying onmodels and inference instead [Wikipedia]
I Design algorithmsI Measure the performance of the algorithm to baselines
I Classification, clustering, forecasting accuracyI Error metrics
![Page 53: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/53.jpg)
Challenges in use of new data sources for machine learning
I Where to get data?
I How to validate your algorithm?
![Page 54: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/54.jpg)
Where to get data?
I Collect data by deploying some sensing technology
(GPStrackers, Wifi scanning, proximity sensing)
I Search for an alternative solution, collect data from a source(Crawl the web, use APIs)
![Page 55: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/55.jpg)
Where to get data?
I Collect data by deploying some sensing technology (GPStrackers, Wifi scanning, proximity sensing)
I Search for an alternative solution, collect data from a source(Crawl the web, use APIs)
![Page 56: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/56.jpg)
Where to get data?
I Collect data by deploying some sensing technology (GPStrackers, Wifi scanning, proximity sensing)
I Search for an alternative solution, collect data from a source(Crawl the web, use APIs)
![Page 57: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/57.jpg)
How to validate your algorithm?
Case: You are designing an algorithm to find periodic patternsfrom people’s trajectory data
The recurring issues of ground-truth
![Page 58: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/58.jpg)
How to validate your algorithm?
Case: You are designing an algorithm to find periodic patternsfrom people’s trajectory dataThe recurring issues of ground-truth
![Page 59: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/59.jpg)
Dealing with ground truth issues
I Ask data collectors to label their data (e.g. Lausanne DataCollection Campaign)
I Ask other people to label your data (e.g. Mechanical Turk,Amazon SageMaker Ground Truth)
I Validation through defining a higher level machine learningtask (e.g. does finding periodic patterns help me predict thefuture of trajectories better?)
I Validation using additional data which is considered highlycorrelated with the pattern you are looking for (people whowork 5 days a week have a strong periodic pattern. Can wedistinguish them better from people who do not work?)
I Synthetic data generator (e.g. data simulated based on knownpatterns)
I Make synthetic data as close as possible to actual data (addnoise, missing data, random patterns,...)
I Mess with data in all possible ways to make sure youralgorithm works all the time
![Page 60: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/60.jpg)
Dealing with ground truth issues
I Ask data collectors to label their data (e.g. Lausanne DataCollection Campaign)
I Ask other people to label your data (e.g. Mechanical Turk,Amazon SageMaker Ground Truth)
I Validation through defining a higher level machine learningtask (e.g. does finding periodic patterns help me predict thefuture of trajectories better?)
I Validation using additional data which is considered highlycorrelated with the pattern you are looking for (people whowork 5 days a week have a strong periodic pattern. Can wedistinguish them better from people who do not work?)
I Synthetic data generator (e.g. data simulated based on knownpatterns)
I Make synthetic data as close as possible to actual data (addnoise, missing data, random patterns,...)
I Mess with data in all possible ways to make sure youralgorithm works all the time
![Page 61: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/61.jpg)
Dealing with ground truth issues
I Ask data collectors to label their data (e.g. Lausanne DataCollection Campaign)
I Ask other people to label your data (e.g. Mechanical Turk,Amazon SageMaker Ground Truth)
I Validation through defining a higher level machine learningtask (e.g. does finding periodic patterns help me predict thefuture of trajectories better?)
I Validation using additional data which is considered highlycorrelated with the pattern you are looking for (people whowork 5 days a week have a strong periodic pattern. Can wedistinguish them better from people who do not work?)
I Synthetic data generator (e.g. data simulated based on knownpatterns)
I Make synthetic data as close as possible to actual data (addnoise, missing data, random patterns,...)
I Mess with data in all possible ways to make sure youralgorithm works all the time
![Page 62: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/62.jpg)
Dealing with ground truth issues
I Ask data collectors to label their data (e.g. Lausanne DataCollection Campaign)
I Ask other people to label your data (e.g. Mechanical Turk,Amazon SageMaker Ground Truth)
I Validation through defining a higher level machine learningtask (e.g. does finding periodic patterns help me predict thefuture of trajectories better?)
I Validation using additional data which is considered highlycorrelated with the pattern you are looking for (people whowork 5 days a week have a strong periodic pattern. Can wedistinguish them better from people who do not work?)
I Synthetic data generator (e.g. data simulated based on knownpatterns)
I Make synthetic data as close as possible to actual data (addnoise, missing data, random patterns,...)
I Mess with data in all possible ways to make sure youralgorithm works all the time
![Page 63: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/63.jpg)
Dealing with ground truth issues
I Ask data collectors to label their data (e.g. Lausanne DataCollection Campaign)
I Ask other people to label your data (e.g. Mechanical Turk,Amazon SageMaker Ground Truth)
I Validation through defining a higher level machine learningtask (e.g. does finding periodic patterns help me predict thefuture of trajectories better?)
I Validation using additional data which is considered highlycorrelated with the pattern you are looking for (people whowork 5 days a week have a strong periodic pattern. Can wedistinguish them better from people who do not work?)
I Synthetic data generator (e.g. data simulated based on knownpatterns)
I Make synthetic data as close as possible to actual data (addnoise, missing data, random patterns,...)
I Mess with data in all possible ways to make sure youralgorithm works all the time
![Page 64: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/64.jpg)
Dealing with ground truth issues
I Ask data collectors to label their data (e.g. Lausanne DataCollection Campaign)
I Ask other people to label your data (e.g. Mechanical Turk,Amazon SageMaker Ground Truth)
I Validation through defining a higher level machine learningtask (e.g. does finding periodic patterns help me predict thefuture of trajectories better?)
I Validation using additional data which is considered highlycorrelated with the pattern you are looking for (people whowork 5 days a week have a strong periodic pattern. Can wedistinguish them better from people who do not work?)
I Synthetic data generator (e.g. data simulated based on knownpatterns)
I Make synthetic data as close as possible to actual data (addnoise, missing data, random patterns,...)
I Mess with data in all possible ways to make sure youralgorithm works all the time
![Page 65: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/65.jpg)
Example generation of synthetic data
10
10[LWH12]
![Page 66: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/66.jpg)
Lessons learned
I Urban Computing field was born due to the adoptionubiquitous and pervasive systems
I Old and new data sourcesI Old → questionnaire, census, survey, ...I New → citizen generated data, open data, digital businesses
data, ...
I Urban Computing design computational methods for analysisof new data sources for old urban problems
I Adoption of machine learning methods for new data sourcesrequires careful experiment design and identifying validationapproaches (ground truth, labeling, connecting data sources,synthetic data)
![Page 67: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/67.jpg)
End of theory!
![Page 68: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/68.jpg)
Part 3: Hands-on lab
![Page 69: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/69.jpg)
References I
Daniel Arribas-Bel, Accidental, open and everywhere:Emerging data sources for the understanding of cities, AppliedGeography 49 (2014), 45 – 53, The New Urban World.
Dani Arribas-Bel, Geographic data science’17, 2018.
Zhenhui Li, Jingjing Wang, and Jiawei Han, Mining eventperiodicity from incomplete observations, in proc. 18th ACMSIGKDD, ACM, 2012, pp. 444–452.
Andreea-Cristina Petre, Cristian Chilipirea, Mitra Baratchi,Ciprian Dobre, and Maarten van Steen, Chapter 14 - wifitracking of pedestrian behavior, Smart Sensors Networks,Intelligent Data-Centric Systems, 2017, pp. 309 – 337.
Eric Paulos and Elizabeth Goodman, The familiar stranger:anxiety, comfort, and play in public places, in proc. SIGCHI,ACM, 2004, pp. 223–230.
![Page 70: Urban Computing › docs › L1 › UC-slides-1.pdfAssignment 1 - Deadline selecting a paper 3 28 Feb 2020 Assignment 1 Spatial data processing - 4 06 Mar 2020 Assignment 1 Spatio-temporal](https://reader033.vdocuments.us/reader033/viewer/2022060501/5f1b83e4ff64051dd03cf8e9/html5/thumbnails/70.jpg)
References II
JH van Staalduinen, J Tetteroo, D Gawehns, and M Baratchi,An intelligent tree planning approach using location-basedsocial networks data.
Yu Zheng, Licia Capra, Ouri Wolfson, and Hai Yang, Urbancomputing: concepts, methodologies, and applications, ACMTIST 5 (2014), no. 3, 38.