COMP20008
Elements of Data Processing
Semester 2 2019
Big Data and Analytics – Cool! But .. do the humans really understand where they are headed?
Dr Suelette Dreyfus
School of Computing and Information Systems
Introduction to your lecturer
Who is Dr Suelette Dreyfus – and why does she know about digital privacy?
State of the Art in Consumer Data Tracking and Its Impact on Consumers in Australia
Report written by Dana McKay, with contributions by Yung Ju Chua, Shanton Chang, Suelette Dreyfus, Monica Whitty, Jeannie Marie Paterson, Pan Zhan, Garreth Hanley and Andrew Clausen
September 2019
Funded by
Embargoed until Friday Nov 8
Analytics is being used to track and analyse individuals
Former US National Security Agency executive Thomas Drake referencing George Orwell’s famous book about state surveillance ‘1984’:
“I don’t want to be Winston cowering in a corner because its only place the cameras couldn’t reach”
Source: https://www.theguardian.com/technology/2019/oct/08/us- whistleblower-thomas-drake-says-speech-was-cancelled-due- to-government-pressure
Old vs New Surveillance
Old surveillance:
• ‘What you say’: Transaction-based
• Conversations
• Postal Mail
• Wire taps on phones
• Chat logs
New surveillance:
‘Who you are’
• movement tracking of
consumers in private-public spaces
• facial recognition
• mood analysis
AND .. New New Surveillance
‘Who we say you are’
1. Interpretationsofa consumer’s
capabilities and self
2. Based on untested, non-transparent ‘science’
3. Potentially used in a discriminatory manner
Biometric Mirror
Biometric Mirror
Analytics being used to track and analyse individuals
Ex: Capable. Not Active? Airplanes:
Carriers w/ installed Panasonic IFE systems:
SQ, AA, EK, QF
• Singapore stated they are ‘disabled’.
• AA: ‘Never been activated’
• UA: ‘Possible future use’
Cameras have become the norm, not the exception, for most hardware
Hacking risk to privacy?
Analytics being used to track and analyse individuals
Capable. And Active?
Emporium shopping centre
(Melbourne)
Analytics being used to track and analyse individuals
Capable. And Active?
Doncaster shopping centre
(Melbourne)
Analytics being used to track and analyse individuals
Capable. And Active.
7-11 Petrol Station
Analytics being used to track and analyse individuals
Spot the Camera at top of Point of Sale
DART
Digital-outdoor Audience in Real Time
• Collection of data combined with responsive advertising
• Real time responsiveness
• “The most intelligent out of home audience
measurement system”
• “So advanced its like having a digital analyst assessing audience while they view your outdoor content in real time.
• “A DA anonymously tracking over 2 m Australians weekly
• “Measures those viewing your content based on Age, Gender, when they viewed, how many viewed across entire length of campaign”
Analytics being used to track and analyse consumers
(Continued)
• Reads audience characteristics with precision
• Identifying 18 demographic profiles
• It even highlights facial features
• “Know viewing habits”
• “Know moods” (eg sentiment analysis)
• “Have your content triggered when the relevant target audience is watching”
• Its not called campaigning but rather:
• “It’s your campaign intelligence” (eg facial slurping and analysis)
Tracking and analysing citizens too..
(Continued)
• Not only the commercial sector
• Governments are putting forward more related laws
• Parliamentary Joint Committee on Intelligence and Security review of Identity-matching Services Bill 2019 and
the Australian Passports Amendment (Identity-matching Services) Bill 2019
• October 2017, the Prime Minister and state and territory leaders agreed to establish a National Facial Biometric Matching Capability & signed Intergovernmental Agreement on Identity Matching Services.
• Easier for law enforcement agencies to identify people
• all jurisdictions will be able to use the new face matching
services to access passport, visa, citizenship, and driver licence
images.
Sources:https://www.aph.gov.au/Parliamentary_Business/Committees/Joint/Intelligence_and_Security /Identity-Matching2019
And https://www.efa.org.au/2019/10/11/scheduled-facial-recognition-public-inquiry-cancelled-by- australian-government/
Models of consent in data gathering from consumers
1. Forced consent: A clickwrap ‘take it or leave it’ approach
2. Unforced consent: Fully informed, and where consumers have some control over their own data
3. No consent: No information or opportunity to opt out is provided to users of a space or service.
To have any control over their data, consumers must be able to give unforced or genuine informed consent.
But with facial ‘slurping’, #3 becomes the norm.
Deeper Issues
1. Do you ‘own’ the exterior manifestation of yourself?
• Can we talk about ‘digital citizenship’ if you have no enforceable claim of rights?
Summary Issues
2. Do we have the right to know how others ‘read’ us in ‘forced consent’ or ‘no
consent’?
• Govt institutions
• Commercial service providers
• Other people (consumers) who use these services (eg dating, housemate finding services)
• Privacy is contextual: one Insta post versus the entire Twitter feed
Summary Issues
2. Do we have the right to correct how others ‘read’ us in ‘forced consent’ or ‘no consent’?
• Govt institutions
• Commercial service providers
• Other people (consumers) who use these services (eg dating, housemate finding services)
•
• •
Summary:
3 big emerging areas of concern
Movement tracking of consumers and citizens in private-public spaces
Facial recognition
Mood / sentiment analysis
Ethical implications for Individuals: Some factors to consider
Data Control
The extent to which an individual is empowered to audit the access to, storage, exploitation and modification of data about the individual.
Awareness
The extent to which an individual is mindful in consenting as to what data is collected about them and how it is used.
Trust
The extent to which an individual can have confidence that the parties who have access to their data respect the individual’s rights.
Privacy
The extent to which an individual is able to restrict the disclosure of their personal information. “If you have something that you don’t want anyone to know, maybe you shouldn’t be doing it in the first place” ?
Choice
The extent to which an individual is able to make choices without being unfairly discriminated against or constrained by the use of big data and analytics.
Anxiety
The extent of psychological discomfort engendered by the collection and use of personal information for big data analytics purposes.
Ethical implications for Organisations
Data Quality
The extent to which organisations ensure the accuracy, currency, completeness and validity of big data.
Data Sourcing
The extent to which organisations collect, buy, and aggregate data from multiple sources in a manner that respects the rights of individuals.
Data Sharing
The extent to which organisations share, sell or otherwise disclose data in a manner that respects the rights of individuals.
Decision Making
The extent to which big data analytics and resulting organisational decisions respect the rights of individuals.
Ethical Culture
The extent to which organisations have values, norms and shared beliefs that promote ethical big data analytics practices through education, training and other means.
Ethical Data Governance
The extent to which organisations articulate ethical standards, decision rights and responsibilities for sourcing, analysing and sharing big data.
Behaviour
The extent to which organisational actors behave consistently with their organisation’s ethical culture and standards.
Reputation
The extent to which relevant stakeholders (e.g. customers) believe an organisation will manage and utilise data about them ethically.
Competitive Pressure
The extent to which organisations are subject to pressure to compete using big data analytics unethically.
Ethical implications for Society
Power Imbalance
The extent to which power in society is imbalanced by the use of big data analytics by a dominant group, organisation or government.
Coercion
The extent to which participation and functioning in society is dependent on contributing one’s own data to a collection for analysis.
Social Awareness
The extent to which members of a society are aware of the role of big data analytics in directing and regulating behaviour in the society.
Surveillance
The extent to which the lives of individuals in a society are observed, monitored, measured and profiled.
Principles and Guidelines
The extent to which effective principles and guidelines exist to protect the rights of individuals impacted by big data analytics.
Authority
The extent to which an entity (e.g., government, professional association) acts to enforce, through sanctions or other means, the rights embodied in the established principles and guidelines for big data analytics.
Social Mindset
The extent to which society collectively feels anxious and oppressed (or opposingly assured and empowered) about the use of big data analytics.
Societal actors need to provide oversight and regulate
EU General Data Protection Regulation (enforced since May 2018) Aims to regulate and protect data privacy for all EU citizens.
• Penalty4%ofannualglobalturnoveroftheorganizations.
The consent
• shouldbeclear,concise,nottoolongandintelligiblywritten—shouldattachthe
reasons of data collection and analyses.
• individualshavetherighttowithdrawtheconsentwiththesameeasinessthat
they have previously agreed with.
Accessing individual’s data
• Individualshavetherighttoaskforacopyoftheirpersonaldatatogetherwith
information regarding the processing and purpose of data collection and
analyses from a controller
• Individualshavetherightofdataportability,whichmeansthattheycan
transfer their data from one controller to another. http://www.eugdpr.org/eugdpr.org.html
Summary: What can we do?
We need to empower individuals
• Educate individuals, raise social awareness
• Provide data access and control (e.g. Google activity)
– http://www.abc.net.au/4corners/stories/2017/04/10/4649443.htm
Define and develop a culture of acceptable data use
• Organizations should internalize the costs
• Genuine consent from individuals
• Be transparent and clearly communicate intent of data collection and analytics/AI
• Adopt rules for responsible big data research
• Provide data control to individuals
Acknowledgements
The Research Project Team
Assoc. Prof. Shanton Chang – CIS, University of Melbourne Prof. Jeannie Marie Paterson – Law, University of Melbourne Prof. Monica Whitty – Arts, University of Melbourne
Dr Andrew Clausen – Economics, University of Edinburgh Dr Dana McKay – CIS, University of Melbourne
Yung Ju Chua – CIS, University of Melbourne
Pan Zhan- ENG, University of Melbourne
Garreth Hanley – Journalist
Dr Suelette Dreyfus – CIS, University of Melbourne
And
The Consumer Policy Research Centre in Melbourne, Victoria
25
References
Anscombe, L. (2017). Westfield is using facial detection software to watch how you shop. Retrieved from https://www.news.com.au/finance/business/retail/westfield-isusing- facial-detection-software-to-watch-how-you- shop/newsstory/7d0653eb21fe1b07be51d508bfe46262
Culnane, C., Rubinstein, B. I. P., & Teague, V. (2017). Health Data in an Open World. arXiv eprints. https://ui.adsabs.harvard.edu/\#abs/2017arXiv171205627C
Chua, YJ (2019) Privacy and discrimination of consumers in a world of ubiquitous computing
26