WWW2010 : danah boyd

danah boyd is with Microsoft Research New England and a research fellow at the Berkman Center for Internet and Society at Harvard University; she was one of the first people to research social technical networks of teens

Privacy and Publicity In The Context of Big Data

Privacy

• What are teens doing
• Technical (phising etc)
• Concerns are not new – long before the Net
• What is new has to do with big data

Big Data

• Remix, aggregate – average people are producing data
• Specifically referring to “social data” – people, their activities, their interactions, their behaviors [FB, Twitter]

Sense-Making

• Data is cheap; analysis is not
• What does it mean to engage with this data? Ethics, privacy, publicity

Methodology

• WWW is one of the places where we see big data being researched
• Ethics : ethnography – map cultures, figure out what people do. I started out in computer science, visualizing social networks. I wanted to understand why.
• Georeg Homans : the mthods of social science are dear in time and money and getting dearer every day
• Vint Cerf: We never, ever in the history of mankind have had access to so much info so quickly and easy

* Bigger data are not always better data
* Not all data are created equal
* What and why are different question
* Be careful of your interpretations

SAMPLING

• Quality more impt than quantiy
• big-ness != whole-ness | Twitter has all Twitter but most researchers only have a sample. If you are trying to understand topicalness, your analysis will be wrong no matter how large your sample is
• know your biases – if you have all tweets and can pull a random sample, it is a random sample of tweets not users. some users have multiple accounts, some consume without accounts, some tweet more than others
• not all data are created equal
• “better” networks
• different kinds of social networks : articulated, behavioral, personal … articulated (people you list off on FB etc — lots of people here who are not your “closest” friends — you can’t FB friends list and say you’ve analyzed a person’s social network), behavioral (same room, email, cellphone) neither are the same as personal networks
• Homopholy has been shown to play out in all of these in interesting ways
• “Tie” strength — the person I list as my top friend may be there because of politics; just because I talk to my collaborators more than my mom does not mean that they
• No one loves big data more than marketers and no one misunderstands big data more than marketers [coke/coca-cola and teens example – linked for different reason than what coca-cola thought]

What and Why are not the same thing!

• Cobot + Lambdamoo
• Frequency is not tie strength (drama over misinterpretation)

Interpretation

• Fallacy that qualitative research is interpretation and quantitative research is about producing facts
• Robin Dunbar’s work -> you could only keep up with gossip of 150 people max. Friendster interpreted this as saying people would have no more than 150 friends and capped it there
• Hardest part of research is interpretation – it’s why social scientists are focused on methodology

The #1 threat to privacy today is our focus on big data

We have perverted the notion of public … [find this quote]

Just because data is accessible doesn’t mean that using it is ethical — privacy is contextual — people trust each other to maintain context and they feel violated when this is broken

“These walls have ears” — Chaucer (1387)

People believe that they understand the context in which they are operating and they get upset when this changes – FB – technology DOES have eyes and mouths

Big data isn’t arbitrary data – it is data about people’s lives — the process of sharing and using

* security through obscurity is reasonable

• how we act in public spaces depends on context
• often ephemeral
• surveillance cameras capture cartwheels
• people change behavior when they know that they are being recorded
• mediated situations like FB but amplifying is different — people are developing skills
• but when we keep changing the context people get confused — people’s encounters with social systems
• civil inattention [get this quote] you may be able to stare at everyone who walks by, but you don’t. why is ok to demand the right to stare at everyone online just because you can

* not all publicly accessible data is meant to be publicized

• using a sense of obscurity for context
• paparazzi make lives of celebrities hell
• when we argue for the right to publicize any data that are successful we are arguing that everyone life the life of a celebrity in this sense
• psychological consequences?
• PII – personally identifiable information
• PEI – personally embarrassing information
• aggregating and distributing data out of context is a privacy violation

* Privacy is not access control

• limiting access can be one tool but it’s not the same as privay

* challenging questions

• “should we?”

* publicity

• hard to distinguish between content meant to be aggregated and that which isn’t (context)

* data are people

FACEBOOK

• Built its reputation on being a closed system
• Interpreted by the public as “anti-myspace” – narrated as closed, intimate – “safer” because it was more private
• First impressions matter
• To this day, many of the average FB users I talk to believe it is about privacy
• Students Against Facebook News Feed -> already accessible but it took implicit content and published as explicit content — changed the context and people wanted to opt-out — it changed peoples’ behaviors — they thought differently about how — those who joined after 2006 have different norms
• Beacon – 2007 advertising messages — made individuals advertisers for their friends — opt in by default — not as visible as news feed – only learned about it when something went wrong — dismantled and settled law suit
• Last year – “invite” to change privacy settings — the default choice was /everyone/ — for most people, they just clicked through. FTC challenged. 65% of users made their content publicly accessible (researchers know most people don’t change defaults)
• I asked people to describe their privacy settings and then we went through their settings. Not a single person I interviewed had a mental map that
• clueless, confused and outright screwed (proverbial boiling frog)
• last week announcement – it relies on the changes from FB (a form of trickery)
• you have to opt-out of each individual partner site on FB and the partner site
• I spent six hours trying to figure out how to turn off like for politicians

CONTEXT

• opt-out is in the better interest of companies not people
• people are engaging with FB for personal reasons — huge ethical issue/challenge

NOT ABOUT HIDING

Regulation – Lessig – Code

• Changes are coming about because of changes in architecture, changes in code
• Technology’s role as regulator has rapidly changed
• Social norms haven’t radically changed
• Law is interesting player in all of this

GET INVOLVED