Analysing Daily Behaviours with Large-Scale Smartphone Data
-
Upload
neal-lathia -
Category
Science
-
view
60 -
download
0
Transcript of Analysing Daily Behaviours with Large-Scale Smartphone Data
![Page 1: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/1.jpg)
Analysing Daily Behaviours with Large-ScaleSmartphone Data
@neal_lathiaComputer LaboratoryUniversity of Cambridge
![Page 2: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/2.jpg)
BackgroundSmartphones as Research ToolsCase 1: Public TransportCase 2: Subjective Wellbeing & BehaviourCase 3: Behavioural InterventionChallenges, Opportunities, Questions
![Page 3: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/3.jpg)
Background
![Page 4: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/4.jpg)
![Page 5: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/5.jpg)
Smartphones as Research Tools
![Page 6: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/6.jpg)
“by 2025, when most of today’s psychology undergraduates will be in their mid-30s, morethan 5 billion people on our planet will be usingultra-broadband, sensor-rich smartphones farbeyond the abilities of today’s iPhones, Androids,and Blackberries.”
Miller
![Page 7: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/7.jpg)
![Page 8: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/8.jpg)
![Page 9: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/9.jpg)
AccelerometerGPS / Wi-FiGyroscopeBluetoothMicrophoneHumidityTemperaturePhone / Text LogsDevice LogsSocial Media APIsApp Usage
![Page 10: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/10.jpg)
Accelerometer | Physical ActivityGPS / Wi-Fi | MobilityGyroscope | OrientationBluetooth | Co-LocationMicrophone | Ambient AudioHumidity | EnvironmentTemperature | EnvironmentPhone / Text Logs | SocialisingDevice Logs | NetworkSocial Media APIs | SocialisingApp Usage | /Information Needs
![Page 11: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/11.jpg)
Case 1: Public Transport
N. Lathia, L. Capra. Tube Star: Crowd-Sourced Experiences on PublicTransport. In 11th International Conference on Mobile and UbiquitousSystems. London, December 2014.
![Page 12: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/12.jpg)
![Page 13: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/13.jpg)
![Page 14: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/14.jpg)
Conclusion: potential for smartphones as a near-real time passenger surveying tool to collect qualitative trip data.
Major Limitation: amount of data received. Still relies on reported behaviour.
![Page 15: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/15.jpg)
![Page 16: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/16.jpg)
Accelerometer (activity)
GPS / Wi-Fi (location)
Gyroscope (orientation)
Bluetooth (co-location)
Microphone (audio context)
Environment (temperature)
Phone / Text Logs (sociability)
Device Logs (e.g., network)
Social Media APIs (crowdsourcing)
App Usage (information seeking)
![Page 17: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/17.jpg)
Method1. Collect Wi-Fi scans that match“Virgin Media Wi-Fi”
2. Manually label “unknown”stations (“Where are you?”)
3. Apply heuristic-based clusteringalgorithm to determine stationvisits, paths, travel times.
Preliminary Data34 users; 234,769 Wi-Fi scans,106,793
![Page 18: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/18.jpg)
Sequences of Wi-Fi Connections
![Page 19: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/19.jpg)
Oyster Card transaction
![Page 20: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/20.jpg)
Journey Planner:Victoria Line to Oxford CircusCentral Line to Mile End27 Minutes
![Page 21: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/21.jpg)
Measured:Victoria Line to VictoriaCircle/District to Mile End29 minutes
21:54:02
21:56:06
21:58:13
22:12:00
22:15:41
22:23:05
![Page 22: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/22.jpg)
Oyster card transaction
![Page 23: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/23.jpg)
Journey Planner:Piccadilly Line to HolbornCentral Line to West Ruislip56 Minutes
![Page 24: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/24.jpg)
Measured:Piccadilly Line to King's CrossVictoria Line to Oxford CircusCentral Line to West Ruislip74 minutes
10:56:06
11:15:39
11:32:55
12:10:09
![Page 25: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/25.jpg)
Capturing Routes: Given an O-D pair, count the % of times that another station appears as an intermediary
Observing Mistakes? E.g., 2.18% of trips from Pimlico to Victoria Station go via Green Park (wrong direction).
Non-Adjacent Pairs of Wi-Fi Connections
![Page 26: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/26.jpg)
![Page 27: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/27.jpg)
40%
10%
10%
5%
![Page 28: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/28.jpg)
![Page 29: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/29.jpg)
5%
41.70%64.7%
![Page 30: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/30.jpg)
Continued (lessened) limitations:Data is not “complete” - phones do not always connect.Data is now “noisy” by capturing route errors, “strange”behaviours.
Direct application:Transport route choices in individuals
With more scale:Granular origin-destination + distributions of route data.
With more data:I.e., precise locations of Wi-Fi hotspots (e.g., platform,entrance)
With more sensors:What actual behaviours are occurring?
![Page 31: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/31.jpg)
Case 2: Subjective Wellbeing & Behaviour
N. Lathia, K. Rachuri, C. Mascolo, P. Rentfrow. Contextual Dissonance:Design Bias in Sensor-Based Experience Sampling Methods. In ACMInternational Joint Conference on Pervasive and Ubiquitous Computing.Zurich, Switzerland. September 2013.
N. Lathia, G. Sandstrom, P. Rentfrow, C. Mascolo. Happy People LiveActive Lives. In prep.
![Page 32: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/32.jpg)
“A sample of 222 undergraduates was screenedfor high happiness using multiple confirmingassessment filters. We compared the upper 10%of consistently very happy people with averageand very unhappy people. The very happy peoplewere highly social, and had stronger romantic andother social relationships than less happygroups...”
Diener, Seligman. Very Happy People. In Psychological Science 13 (1). Jan 2002.
![Page 33: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/33.jpg)
![Page 34: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/34.jpg)
![Page 35: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/35.jpg)
![Page 36: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/36.jpg)
![Page 37: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/37.jpg)
angry anxious lonely
relaxedenthusiasticcalm
![Page 38: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/38.jpg)
![Page 39: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/39.jpg)
![Page 40: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/40.jpg)
Hemminki, Nurmi, Tarkoma. Accelerometer-Based Transportation Mode Detection onSmartphones. In ACM Sensys 2013.
Statistical: mean, standard deviation, median, etc.Time: auto-correlation, mean-crossing rate, etc.Frequency: FFT, spectral energy, etc.Peak: volume, intensity, skewness, etc.Segment: e.g., velocity change rate
![Page 41: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/41.jpg)
Example: 85 Users
![Page 42: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/42.jpg)
![Page 43: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/43.jpg)
![Page 44: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/44.jpg)
![Page 45: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/45.jpg)
![Page 46: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/46.jpg)
![Page 47: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/47.jpg)
Case 3: Behavioural Intervention
![Page 48: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/48.jpg)
![Page 49: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/49.jpg)
![Page 50: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/50.jpg)
![Page 51: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/51.jpg)
Challenges & Questions
![Page 52: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/52.jpg)
1. Software Engineering / Expectations2. Marketing3. Control over target population4. Understanding sensor data5. Writing code6. Finding research value
![Page 53: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/53.jpg)
1. Blurred lines between research and practice2. High potential for multi-disciplinary impact3. Cheap to roll-out to huge audiences4. Accessible to 'everyone'5. Wearables are coming!
![Page 54: Analysing Daily Behaviours with Large-Scale Smartphone Data](https://reader031.fdocuments.in/reader031/viewer/2022032217/55a786251a28ab8d188b462f/html5/thumbnails/54.jpg)
Can I run a study likeEmotion Sense?
Yes, with Easy M. Ageneralised sensor-enhanced experiencesampling tool.