Here are some quotations from latanya sweeneys paper, that tim oreilly appeared unaware of. Named for the united states navy rear admiral who was a trailblazing female. Data privacy in the age of big data towards data science. Anonymous data cannot be manipulated to reidentify individuals, whereas deidentified data can be. For consumers, an open data society is a misnomer the. Her vision and analytic approach to this emerging field have created a welldeserved. We present a computer program named data y that maintains anonymity in medical data by automatically.
This article is more than 9 years old buying back your own privacy. Sep 26, 2011 the latanya sweeney result was the first to show that once you can mix and match data sets, pii is just not enough to provide privacy. One thing that comes to mind here is latanya sweeneys data map. Our work fits within this general class of solutions that seeks to redact parts of images for. We thank the ftc for allowing the information to be reported publicly. Figure 1 below is a simple venn diagram with two intersecting circles. In november 2017, the running app strava released a data visualization map showing every. And nowadays, of course, data mining multiple data sets is big business. Sweeney l, yoo j, perovich l, boronow k, brown p, brody j. They are then forced to choose an access policy for their data. If you need to print pages from this book, we recommend downloading it as a pdf. Latanya sweeney explains why tech companies are so powerful.
When used for advertising, they can reproduce our own prejudiced. Nov 05, 20 the researchers analyzed the mobile phone data together with a simple malaria transmission model based on infection prevalence data, and in doing so were able to map routes of malaria parasite dispersal. The latanya sweeney result was the first to show that once you can mix and match data sets, pii is just not enough to provide privacy. In fact, a few companies are challenging the norm of corporate data hoarding by. Computational disclosure control a primer on data privacy. Latanya sweeney explains why tech companies are so. Among these, they found that 33 states collected and shared hospital discharge data publicly.
The first type of protection system, adopted in a wide range of communities and environments, is based on deidentification deid. The re identification of governor william welds medical. May 26, 20 the fact that i am producing data and companies are collecting it to monetize it, if i cant get a copy myself, i do consider it unfair, says latanya sweeney, the director of the data. Latanya sweeney also addresses this point in her phd thesis. Latanya arvette sweeney is a professor of government and technology. Since then, the anecdote is now repeated over and over, as anonymized datasets with. Ethnicity, visit date, diagnosis, procedure, medication, total charge. I formally define and present null map, k map and wrong map as models of protection. Increasingly you might come across an interesting set of interactive charts from a public body, or an interactive map. Latanya sweeney, as chief technology officer at the u.
Latanya arvette sweeney is a professor of government and technology in residence at. Click on a circle above for names of organizations and details of data shared. She was easily able to identify the state governors. Purloined information could serve as an answer key to solve the deidentification puzzle, since information in stolen medical data and anonymized patient dossiers may detail the same treatment by a named doctor with certain procedure codes, times of treatment, and prescribed drugs.
Jan 01, 2005 the sharing and application of personspecific genomic data pose complex privacy issues and are considered the foremost challenges to the biomedical community. Re identification of governor william weld 4 cambridge and likely the subject of ballet casting photoops, he was also certain to be listed in the cambridge voter registration rolls and so another key hurdle required for a reidentification attempt using a voter list attack was overcome. For example, when sweeney reidentified hospital discharge data released by washington state, her reidentification exposed records that included sensitive information such as references to. Sharing of sensitive data by android apps left to domains right. The solution provided in this paper includes a formal protection model named kanonymity and a set of accompanying policies for deployment.
Privacy preservation techniques in big data analytics. Mobile data for development primer linkedin slideshare. I begin by demonstrating ways to learn information about entities from publicly available information. The thinning line between commercial and government surveillance. A dataset is kanonymous if you cannot distinguish any one record from k1 other records based on identifiers or quasiidentifiers. We mainly work with free data from the openstreetmap project and use the lean openstreetmap tools for cartography of all kind. May 17, 2016 to avoid biasing, there was no overlap between data sets of different sizes. Sweeneys work has focused on identifying those unexpected places and on showing that its possible to determine some peoples identities from medical data, even after the records have been stripped of personal information. Select an app below for the names of domains the app sent data to and the contents sent. Incredible amounts of data is being generated by various organizations like hospitals, banks, ecommerce, retail and supply chain, etc. Predictably, participants fared better with map based representations, correctly. The following exercise can help cities map data collection to impacts both positive and negative in order. Other work studies different forms of face deidentification 5, 14, 25,38 for privacy protection. Latanya sweeney, then a mit graduate student in computer science.
A survey of behind the scenes personal data sharing to third parties by mobile. In many cases you dont need to do any scraping you just need to know where to look. May 09, 2016 danoff dean of harvard college rakesh khurana announced the appointment of the new faculty deans of currier house. In this age of everincreasing data, it is becoming increasingly. Latanya sweeney, a leading expert in data privacy and anonymization, provided a 67page report and testified for the state bar about the risks to. Latanya sweeney, federal trade commission and harvard university. Reidentification of individuals in chicagos homicide database. Data visualization and plotting data on historical maps 0 data processing. A read is counted each time someone views a publication summary such as the title, abstract, and list of authors, clicks on a figure, or views or downloads the fulltext. I then provide a formal framework for reasoning about disclosure control and the ability to infer the identities of entities contained within the data. To avoid biasing, there was no overlap between data sets of different sizes. These recipients may have no duty to the data subject and no direct relationship or obligation to the. The reidentification of governor william welds medical. The high cost of keeping your personal information personal.
Todays globally networked society places great demand on the dissemination and sharing of personspecific. Ap for most people, every five minute stroll on the web, the. We need to be forward lookinglets think about whats coming during the next 3, 5, or 10 years to address potential threats. We need to be forward lookinglets think about whats coming during the next 3, 5, or 10 years to address. So, kanonymity provides privacy protection by guaranteeing that each released record will relate to at least k individuals even if the records are directly linked to external information. Apr 06, 2011 this article is more than 9 years old buying back your own privacy. The thinning line between commercial and government. In 2001, sweeney became director and founder of the data privacy lab.
For consumers, an open data society is a misnomer the new. Minutes of the technical expert panel meeting aspe. Professor latanya sweeney and sylvia barrett will take their posts this fall. Deanonymizing south korean resident registration numbers shared in prescription data. Predictably, participants fared better with map based representations, correctly identifying twitter users homes roughly 65 percent of the time and their workplaces at closer to 70 percent. Sweeneys groundbreaking research in data privacy has been featured in consumer reports, newsweek, newsday, business week, and the wall street journal, as well as on the television news magazine 2020. Algorithms are great and all, but they can also ruin. In fact, a few companies are challenging the norm of corporate data hoarding by actually sharing some information with the customers who generate it and offering tools to put it to use. Evaluation of the current state of genomic data privacy. Reidentification of individuals in chicagos homicide. Sweeney received her bachelors degree in computer science from harvard, and masters and phd degrees in computer science from massachusetts institute of technology. Decisions are made at the eld and record level at the time of database access, so. In fact, as an abc investigation reported last fall, millions of records can be. We present a computer program named data y that maintains anonymity in medical data by automatically generalizing, substituting, inserting and removing information as appropriate without losing many of the details found within the data.
May 10, 2017 increasingly you might come across an interesting set of interactive charts from a public body, or an interactive map, and you want to grab the data behind it in order to ask further questions. A release provides kanonymity protection if the information for each person contained in the release cannot be distinguished from at least k1 individuals whose information also appears in the release. In this episode carnegie mellon university computer scientist latanya sweeney talks about the changes in privacy due to data collection and approaches to protect privacy in. Since then, the anecdote is now repeated over and over, as anonymized datasets with any personal.
I formally define and present null map, k map and wrong map. More generally, sweeney used 1990 census data to estimate that 0. Open data privacy harvards dash harvard university. Danoff dean of harvard college rakesh khurana announced today the appointment of the new faculty deans of currier house. Federal trade commission, led a group of summer research fellows, jinyan zang, krysta dummit, james graves, and paul lisker, at the ftc in a project to survey data sharing from 110 popular free apps. The 16th grace hopper celebration of women in computing kicks off wednesday in houston. Here are some quotations from latanya sweeney s paper, that tim oreilly appeared unaware of. She was easily able to identify the state governors own records based on a combination of zip code, date of birth and gender. Protecting location privacy in augmented reality using k. Professor of government and technology in residence. We would like to see people have access to all of the data that they produce.
We all leave a trail of personal digital exhaust intentionally and unintentionally, and various. Latanya sweeney, a leading expert in data privacy and anonymization, provided a 67page report and testified for the state bar about the risks to bar candidates privacy in releasing data in the manner proposed by petitioners. Citeseerx document details isaac councill, lee giles, pradeep teregowda. Latanya sweeney, merce crosas, michael barsinai, david obrien, alexandra wood and urs gasser berkman center, steve chong seas, and micah altman mit. May 15, 2017 the data that tracks our behavior feeds into machinelearning algorithms that make judgments about us. Unfortunately, this book cant be printed from the openbook. Datatags will provide dataset owners with simple data handling prescriptions that comply with the numerous regulations that apply to datasets, as well as with data use agreements that. Grace hopper conference spotlights women in tech stem. Sep 22, 2018 incredible amounts of data is being generated by various organizations like hospitals, banks, ecommerce, retail and supply chain, etc. Aug 01, 2007 in this episode carnegie mellon university computer scientist latanya sweeney talks about the changes in privacy due to data collection and approaches to protect privacy in the future, with. As professor of government and technology in residence at harvard university, my mission is create and use technology to assess and solve societal, political and. Personal health data can now be sent in an instant to growing numbers of people and organizations. Tons of data is generated every minute by social media and smart phones.
208 177 1316 1378 208 1404 40 930 1167 237 1460 1045 1376 620 289 251 426 1021 883 753 1638 1622 1498 506 1247 925 468 1211 1300 224 1669 343 381 8 703 977 477 968 355 335 669 1061 337 105 636 1435 1095 110 928 926