A Face Is Exposed for AOL Searcher No

By Massimo Barbaro. 2006

823 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Social Data: Biases, Methodological Pitfalls, and Ethical Boundaries (2019, background): “Prominent releases of anonymized data, such as for the Netflix prize and the AOL’s Search History Database, were later found to provide inadequate protection to users (Barbaro et al., 2006; Narayanan and Shmatikov, 2008).”
Composition attacks and auxiliary information in data privacy (2008, background): “At the same time, the release of information from such repositories can be devastating to the privacy of individuals or organizations [5].”
Gradient-Leaks: Understanding and Controlling Deanonymization in Federated Learning (2018, background): “Consequently, in the aforementioned instances, many identities of participating individuals were re-identiﬁed using a public voter database [75], IMDb movie ratings [57], and search keywords [8] respectively.”
Ten simple rules for responsible big data research (2017, background): “There are numerous examples of researchers with good intentions and seemingly good methods failing to anonymize data sufficiently to prevent the later identification of specific individuals [17]; in other cases, these efforts were extremely superficial [18, 19].”
Privacy in Search Logs (2009, background): “Thelma Arnold from Lilburn, Georgia as searcher number 4417749 [3] in the AOL search log; her queries contained not only enough information to identify herself, but also showed her love for dogs and her friends’ medical ailments.”
Test Collection Based Evaluation of Information Retrieval Systems (2010, methodology): “Although the logs were anonymized, sufficient session information was preserved to allow some people to be identified [21].”
Verifiable differential privacy (2015, background): “…who contributed the data is clearly important, but doing so e ﬀ ectively is highly nontrivial: experience with cases like the AOL search data [1] or the Net-ﬂix prize [3] has shown that even sanitized or aggregate data will often reveal more information than intended, especially in the…”
Practical Secure Aggregation for Privacy-Preserving Machine Learning (2017, background): “…some applications, for other applications stronger guarantees may be needed, as indicated by the failures of ad-hoc anonymization techniques [6], [7], [8], and by the demonstrated capability to extract information about individual training data from fully-trained models (which are essentially…”
Relationships and data sanitization: a study in scarlet (2010, background): “Previous work in sanitization often views a system as a closed world—ignoring outside knowledge that may be available to attackers [7,11].”
Understanding User Behavior Through Log Data and Analysis (2014, background): “Two days later a New York Times story revealed that AnonID 4417749 was Thelma Arnold, a 62 year old woman from Lilburn GA. (Barbaro & Zeller, 2006 ).”
On the tradeoff between privacy and utility in data publishing (2009, background): “This is clearly illustrated when New York Times reporters identified a single user in the search logs published by AOL, causing AOL to remove the data immediately and fire two employees involved in publishing the data [3].”
GUPT: privacy preserving data analysis made easy (2012, unknown): “Examples include the deanonymization of AOL search logs [3], the identiﬁcation of patients in a Massachusetts hospital by combining the public voters list with the hospital’s anonymized discharge list [23] and the identiﬁcation of the users in an anonymized Netﬂix prize data using an IMDB movie…”
Engineering Privacy (2009, background): “When AOL released logs of search queries identified only by pseudonyms, some users were identified because their names or contact information appeared in some of their search queries [92].”