Overview
Medical Informatics
Scrub
Datafly
Genomic identifiability
Patient-centered management
Surveillance
Selective-revelation
Risk assessment server
PrivaMix
Biometrics
Contactless capture
Policy and Law
Identifiability of de-identified data
HIPAA assessments
Privacy-preserving surveillance
Quantitative assessments
A sample of tenured computer science professors in the School of Computer Science at Carnegie Mellon University was asked what would be a compelling measurement for academic work comparable to Dr. Sweeney's. The questions below summarize their responses.
- Is there a significant number of citations to the work?
- Have others been able to successfully build on the work?
- Are any of the newer publications likely to have significant citation counts?
- Is the work appearing in leading publications?
- Have others explicitly recognized the work?
- Has the work had impact in the real-world?
- Has the work contributed to educating the public?
- Has there been funding for or business interest in the work?
No one measure can answer all questions, especially with work that is cross-disciplinary. So, this page uses an array of measurements, one per question, and reports significant results when applying these questions to the work of Latanya Sweeney. 12 Figure 2 provides a summary. A shaded cell indicates a significant finding. Dr. Sweeney's oldest scientific works in database security and medical informatics show statistical significance in scientific citation count measures. Her newer works on face deidentification and contactless fingerprint capture show significance in other scientific measures. Her policy and law and public education work show real-world impact by societal measures. One surprise is that her societal work on the identifiability of real-world data and on HIPAA privacy risk assessments have significant scientific citations. Similarly, her work on surveillance has evidence of real-world impact.
 |
Figure 2. Summary. 15 foci of work, distributed over 7 areas, 5 scientific and 2 societal are represented as rows. The columns are 8 ways of measuring impact corresponding to the 8 questions described above. See links on upper left panel of this page for details for any particular area or work foci.
Below are findings that serve as the basis for shaded cells in Figure 2.
1. Is there a significant number of citations to her work?
- Dr. Sweeney's most cited computer science papers
([cite],
[cite],
[cite],
[cite],
[cite])
have the second
highest citation count among those of Associate Professors (tenured and
nontenured) in the School of Computer Science at Carnegie Mellon
12
and the count is statistically significant at the 99.9th percentile.
- Two of Dr. Sweeney's 7 American Medical Informatics papers
([cite],
[cite])
are among the 15 most cited American Medical Informatics papers to date and have statistically
significant citation counts at the 99.9th percentile.
- The paper of Dr. Sweeney and her student, Brad Malin, on genomic privacy
[cite]
is among the 50 most cited American Medical Informatics papers to date. It's citation count is statistically
significant at the 90th percentile.
- Dr. Sweeney's paper in the Journal of Law, Medicine and Ethics (JLME)
[cite]
was cited in the commentary of the HIPAA Privacy Rule48 has a citation count statistical significant at the
99.9th percentile among medical informatics papers and among JLME papers.
- Dr. Sweeney's uniqueness paper [cite], which reports on the identifiability of basic demographics in the United States, and includes the popular statistic that "87% of the US population are uniquely identified by {date of birth, gender, ZIP}," has a statistically significant citation (at 98.6th percentile) among medical informatics papers.
2. Have others been able to successfully build on my work?
- Three of Dr. Sweeney's computer science papers
([cite],
[cite],
[cite])
are among 5% (58/1156) of the
papers from Associate Professors (tenured and nontenured) in the School of
Computer Science at Carnegie Mellon that enabled successful work by others, as
demonstrated by having at least 5 referencing papers with more citations than the
originating paper.
- Dr. Sweeney's Datafly papers ([cite], [cite]) are among 1% of the American Medical Informatics papers that enabled successful work by others, as demonstrated by having at least 5 referencing papers with more citations than the originating paper.
3. Are any of her newer publications likely to have significant citation counts?
- Dr. Sweeney's most cited papers
([cite],
[cite],
and
[cite])
have the most citations in a year and
seem to still be rising in about half as many years as the most cited papers from
Associate Professors (tenured and nontenured) in the School of Computer Science at
Carnegie Mellon.
- The most popular trend in yearly citation counts among the most cited papers of each
Associate Professor in the School of Computer Science at Carnegie Mellon University is
best modeled by Dr. Sweeney's most cited paper
[cite].
Nineteen other papers share its general trend
of non-decreasing yearly citations (Spearman rank correlation coefficients are significant
at the 99.9% confidence level).
- Dr. Sweeney's recent IEEE Transactions paper that introduces face de-identification [cite] follows the citation trend modeled by her most cited k-anonymity paper [cite] (with Spearman rank correlation rs=1.00, which is statistically significant at the 99.9% confidence level). Using this correlation, a 4 year estimate predicts a citation count, statistically significant at the 99.9th percentile, among those of Associate Professors in the School of Computer Science at Carnegie Mellon University.
4. Is her work appearing in leading publications?
- Two of Dr. Sweeney's papers
([cite]
and
[cite])
appear in IEEE Transactions on Knowledge and
Data Engineering (z-test: among the top 20% of computer science journals).
- Three of Dr. Sweeney's conference papers ([cite], [cite], and [cite]) appear in elite conferences (acceptance rates less than 13%).
5. Have others explicitly recognized her work?
- 5 Career Achievement Awards:
- Appointment, Federal HIT Policy Committee, Privacy and Security Seat.
- Elected Fellow, American College of Medical Informatics (distinguished).
- Privacy Advocate Award, American Psychiatric Association (rare).
- Privacy Leadership Award, Blue Cross Blue Shield Association (rare).
- Scientific American, Featured In-depth Profile (rare).
- 4 Best Paper Awards [cite], [cite], [cite], [cite].
6. Has her work had impact in the real-world?
- Dr. Sweeney's work on the identifiability of data is explicitly cited in the federal commentary
of the HIPAA Privacy Rule, in 4 state supreme or district court decisions, and in 19
law review journal articles.
- Dr. Sweeney has testified or participated in 12 regulatory or advisory committees or briefings for the European Union Commission, DHS, DOD, NCVHS, HCFA, and the U.S. Senate. Topics include privacy-preserving surveillance, data sharing under HIPAA, and medical and genomic privacy.
7. Has her work contributed to educating the public?
- More than 300 news articles cite Dr. Sweeney's work, with 166 in noted news venues. [Source:
Google news and Lexus-Nexus.]
- Dr. Sweeney's work is the specific subject of 28 news articles. Venues include Scientific American, Computerworld, CBS News, ABC News, Newsweek, USA Today, and National Public Radio. Topics include Identity Angel, SSNwatch, privacy-preserving surveillance, HIPAA risk assessments, CameraWatch, patientcentered management, and contactless fingerprint capture.
8. Has there been funding for or business interest in her work?
- Dr. Sweeney received $2.5mil in her own major research grants and worked with others to
receive an additional $15mil. Sources include: DOD, NIH, DOJ, and NSF. Projects
include: Selective Revelation, privacy-preserving surveillance, genomic privacy,
identifiability of data, contactless fingerprint capture and Technology Dialectics.
- Dr. Sweeney had more than 8 Lab Partnerships with real-world stakeholders for research related
to the identifiability of data, patient-centered management, HIPAA assessments, and the
Risk Assessment Server.
- Dr. Sweeney has 5 patent applications in various stages of review, with two patents issued.
These relate to k-anonymity, Datafly, Privamix, and contactless fingerprint capture.
- Two businesses have licenses to Dr. Sweeney's technology to perform HIPAA Risk Assessments.
 |
Figure 3. Citations per year for each of the 5 most cited papers of each Associate Professor in the School of Computer Science at Carnegie Mellon University and for each of Dr. Sweeney's 5 most cited papers. Her 3 most cited papers (LS[46], LS[15], and LS[16]) are among the papers having the most number of citations in a year (tallest peaks). The 3 most cited papers (T1, T2, T3) of the Associate Professor includes two of the the oldest papers (T1 and T2). The citation trajectory labeled M is the single paper having the largest number of citations.
Notes
| 12 | See privacy.cs.cmu.edu/buffer/AssessmentNotebook.pdf for details on all these findings. |
Related links: