Talk:Aadhaar

Unique Identification Authority of India

http://uidai.gov.in/

Severe Bias in Risk Section

The risk section has a clear biased with unsubstantiated and/or misunderstood conclusions. The only supporting references are from the author of the section. For instance, the sentence "There have been concerns expressed in some quarters that the biometrics data is not reliable enough" is only supported by the author's own blog! The argument in the blog is flawed and shows no understanding of biometric accuracy measurements or how these apply to UID's intended use for biometrics. Specifically the author of the blog does not understand that the false positive identification rate is a measure of identification (1 to many) and not verification (1 to 1) and many incorrect conclusions are drawn from this misunderstanding.

The next paragraph in the Risk section goes on to say "level of uncertainty in biometrics is so great that tests tell you nothing". Again this is the interpretation of the author, but it is hard to find any quote remotely similar to this in the papers. This is only one person's opinion and is a gross misrepresentation what was published.

The third paragraph has nothing to do with UID risks. It seems to be there only to push the authors agenda on his personal blog.

Perhaps only the fourth paragraph is unbiased as it points out an issue with the system.

I would recommend deleting the majority of this section unless the author can provide some unbiased support for his conclusions. lift (talk) 16:04, 29 September 2011 (UTC)

Response to Severe Bias in Risk Section

The author is aware of the dangers of bias. It is for that reason that he cites supporting evidence for most statements of fact and in addition includes on his website a folder of evidence of the unreliability of mass consumer biometrics. The Talk contributor @ 16:04, 29 September 2011 may note therein articles published by respectable newspapers and magazines with titles like Security fear over airport face scanners and ID card 'will drown in a billion mismatches' and German Federal Police questions reliability of facial recognition and ID technology leaving passengers waiting etc ...

The author does understand the difference between 1:n and 1:1 and would appreciate it if the Talk contributor @ 16:04, 29 September 2011 would identify, say, three of the mistakes he/she claims that the author has made.

Far from being a misrepresentation, the tests-tell-you-nothing quotation is an entirely accurate inference from the paper written by Messrs Wayman, Possolo and Mansfield, a paper which the Talk contributor @ 16:04, 29 September 2011 might care to read. He will then find this, for example:

There is nothing in a technology test that can validate the simulated data as a proxy for the “real world”, beyond a comparison to the real world data actually available. In other words, technology testing on simulated data cannot logically serve as a proxy for software performance over large, unseen, operational datasets.

and this:

... the test repeatability and reproducibility observed in technology tests are lost in scenario testing due to the loss of statistical control over a wide range of influence quantities.

and this:

Our inability to apply concepts of statistical control to any or all of these factors will increase the level of uncertainty in our results and translate to loss of both repeatability and reproducibility.

and this:

Test data from scenario evaluations should not be used as input to mathematical models of operational environments that require high levels of certainty for validity.

It is demonstrably not hard "to find any quote remotely similar to this in the papers", as the Talk contributor @ 16:04, 29 September 2011 says and the author can see no reason yet to delete anything.

Brother Big (talk) 01:01, 7 October 2011 (UTC) Brother Big (talk) 00:58, 7 October 2011 (UTC) Brother Big (talk) 00:56, 7 October 2011 (UTC) Brother Big (talk) 00:54, 7 October 2011 (UTC) Brother Big (talk) 00:48, 7 October 2011 (UTC)

— Preceding unsigned comment added by David Moss (talk • contribs) 00:35, 7 October 2011 (UTC)

Unresolved issues raised by NIST

The Wikipedia article on Aadhaar which used to appear at http://en.wikipedia.org/wiki/AADHAAR has recently been deleted and that address now redirects to [[1]].

Some issues were raised on the Aadhaar talk page, [[2]], principally by the US National Institute of Standards and Technology (NIST). When the article was deleted, the issues were unresolved. There has been no submission to the talk page since 2 June 2011 at 08:50.

The same issues arise in the case of this article on the Unique Identification Authority of India. Not least because the same text appears, in the Risks section. And so they still need to be resolved.

The Risks section text was accused of being opinion stated as fact. I am the author of that text. I believe I am stating fact. I repeat my request to have examples given to me of opinions being stated by me as fact. I am naturally keen to make any agreed corrections necessary. But i can't just guess what those corrections should be. I would welcome the involvement of NIST on this talk page.Brother Big (talk) 19:04, 9 June 2011 (UTC)

Copy of the Aadhaar talk page

Response to Opinion stated as fact The paper Throwing the towel in quotes extensively from the paper by Messrs Wayman, Possolo and Mansfield attempting thereby at each stage to support its argument. Would the contributor at 65.216.138.253 please list the statements he/she finds "baseless" and I shall be happy then to respond further. Brother Big (talk) 15:38, 27 May 2011 (UTC) ———————————————————————————————————————————————————————————————————————————————

The comments made by NIST have been answered elsewhere as follows:

QUOTE

NIST say: “This research paper described above with a NIST coauthor is about how to apply uncertainty measurements to the testing of biometric technologies”. True. And what the paper says is that the uncertainty cannot be measured, there is no point guessing, the biometrics emperor has no clothes.

NIST say: “Such testing has substantially improved the accuracy of biometric technologies over the years”. False. This is the opposite of what the paper says, which is that it is impossible to measure the accuracy of biometric technologies, they are out of statistical control.

NIST say: “The paper discusses how to understand measurement uncertainties so that law enforcement agencies and other users of biometrics can accurately determine the limitations of their systems”. False, for the same reason. The paper identifies 15 or so influencing factors which contribute to the uncertainty of biometrics, there may be more factors, research is continuing, but, for the moment, measurement of the achievements or limitations of biometrics is impossible.

NIST say: “The paper does not say that biometrics are not useful in real world settings”. True, if we all accept that we know what is meant here by “useful”. That is one of many things that the paper does not say. What it does say is that technology tests and scenario tests tell you nothing about how well biometrics will perform in the field and that operational tests measure nothing. It is quite possible to agree with this statement of NIST’s and yet still believe that Aadhaar, on its own terms, is bound to fail, it cannot identify people uniquely and it cannot verify their identity reliably.

NIST say: “It advocates moving beyond yes-and-no answers in the biometrics area to do a better job of carefully understanding and documenting the variables that affect the accuracy of these technologies as they are used in practice”. True. More research is needed and I, for one, heartily support that. But in advance of that research bearing fruit, what the paper says is that no trial results can provide a valid reason to invest in biometrics. In that case the investment in biometrics is unbusinesslike, irresponsible, unscientific and illogical. Instead of being a reasoned investment, it has more the character of an impulse purchase. Governments are not paid to make impulse purchases using taxpayers’ money.

UNQUOTE

The paper by Messrs Wayman, Possolo and Mansfield states that it is impossible to measure the performance of biometric systems. It follows that it is impossible to say if their accuracy is improving. The authors give as an example of the problem the certificates issued by NIST connection with the USA PATRIOT Act. These certificates state that:

"For purpose of NIST PATRIOT Act certification this test certifies the accuracy of the participating systems on the datasets used in the test. This evaluation does not certify that any of the systems tested meet the requirements of any specific government application. This would require that factors not included in this test such as image quality, dataset size, cost, and required response time be included."

If the best that NIST can do is to say that the test results are the test results but there is no way of knowing how the "participating systems" will perform in the field, that seems to prove the point. Brother Big (talk) 15:57, 27 May 2011 (UTC) ———————————————————————————————————————————————————————————————————————————————

I can see no evidence for the claim made by the user at 65.216.138.253 that "One of the authors (Wayman) is a consultant to UIDAI". Can you provide some? Brother Big (talk) 16:21, 27 May 2011 (UTC)

I see that someone logged on from 65.216.138.253 yesterday at 14:13 and edited this page but didn't answer the question – is there any evidence to support the claim that Mr Wayman is a consultant to UIDAI? It might help if we could have an answer from this IP address. Brother Big (talk) 08:50, 2 June 2011 (UTC)

Wayman was only involved in UIDAI as a presenter at a workshop and conversation he was not much of a participant — Preceding unsigned comment added by 108.28.252.244 (talk) 21:58, 13 December 2012 (UTC) ————————————————————————————————————————————————————————————————————————————————

To the Wikipedia moderators/editors, please advise what I should do. I understand Gail Porter's conflict of interest problem. But I can't just guess what amendments are sought. Here are some of the facts stated as facts in my paper, quoted from the Wayman/Possolo/Mansfield paper:

Is there any hope of inductively extending the results of our technical test more broadly to any other algorithms or databases? A Type B systematic uncertainty evaluation after consideration of changes in the unit of empirical significance and statistical controls over its tangible elements might be of value, provided that the specifics of the changes could be given, but we should not sanctify such a “guesstimate” in an emperor’s cloak of imagined analytic rigor.

... technology testing on artificial or simulated databases tells us only about the performance of a software package on that data. There is nothing in a technology test that can validate the simulated data as a proxy for the “real world”, beyond a comparison to the real world data actually available. In other words, technology testing on simulated data cannot logically serve as a proxy for software performance over large, unseen, operational datasets.

We lack metrics for assessing the expected variability of these quantities between tests and [we lack] models for converting that variability to uncertainty in measurands [the quantities intended here are false positives and negatives, failure to acquire and enrol, and throughput].

... each specific recognition technology (iris, face, voice, fingerprint, hand, etc.) will have specific factors that must be within a state of statistical control. This list of factors is not well understood, although ample work in this area is continuing. For example, recent analysis of iris and face recognition test results shows us that to report false match and false non-match performance metrics for such systems without reporting on the percentage of data subjects wearing contact lenses, the period of time between collection of the compared image sets, the commercial systems used in the collection process, pupil dilation, and lighting direction is to report “nothing at all”. Our reported measurements cannot be expected to be repeatable or reproducible without knowledge and control of these factors.

... the test repeatability and reproducibility observed in technology tests are lost in scenario testing due to the loss of statistical control over a wide range of influence quantities.

... Our inability to apply concepts of statistical control to any or all of these factors will increase the level of uncertainty in our results and translate to loss of both repeatability and reproducibility.

... Test data from scenario evaluations should not be used as input to mathematical models of operational environments that require high levels of certainty for validity.

We can conclude that the three types of tests are measuring incommensurate quantities and therefore [we] should not be at all surprised when the values for the same technologies vary widely and unpredictably over the three types of tests. Brother Big (talk) 16:50, 27 May 2011 (UTC) Brother Big (talk) 16:52, 27 May 2011 (UTC) Brother Big (talk) 01:10, 28 May 2011 (UTC) Brother Big (talk) 01:12, 28 May 2011 (UTC) Brother Big (talk) 01:13, 28 May 2011 (UTC)

———————————————————————————————————————————————————————————————————————————————

Opinion stated as fact

The discussion regarding the paper "Fundamental issues in biometric performance testing" is a misreading of the cited paper, and states conclusions that do not appear in that paper. One of the authors (Wayman) is a consultant to UIDAI. The statement as written is baseless, and appears to be written as a faulty basis for opinion. — Preceding unsigned comment added by 65.216.138.253 (talk) 21:15, 25 May 2011 (UTC)

{{Request edit}} This article does not follow Wikipedia editorial policy. It describes a research paper in a non-neutral, inaccurate way. The March 2010 research paper linked within the Wikipedia article is about how to apply uncertainty measurements to the testing of biometric technologies. Such testing has substantially improved the accuracy of biometric technologies over the years. The paper discusses how to understand measurement uncertainties so that law enforcement agencies and other users of biometrics can accurately determine the limitations of their systems.

The paper does not say that biometrics are not useful in real world settings. It advocates moving beyond yes-and-no answers in the biometrics area to do a better job of carefully understanding and documenting the variables that affect the accuracy of these technologies as they are used in practice.

Moreover, the comment in the Wikipedia article referencing Antonio Possolo of the National Institute of Standards and Technology violates the Wikipedia policy on Biographies of living persons. It attributes statements to Dr. Possolo and/or NIST that are not sourced. {{Request edit}}

Gailjporter333 (talk) 13:50, 26 May 2011 (UTC) Gail Porter Director of Public Affairs National Institute of Standards and Technology

I've cancelled out the 'requested edit' template, because it's quite clear this is not a case of just "change X to Y per Z ref" - I hope discussion can continue (here), and resolution can be found; if you need a specific edit, please ask for that showing consensus. Thanks,  Chzz  ►  02:40, 17 September 2011 (UTC)

NPOV

This page needs a rewrite. The random commentary in the criticism section outweighs the actual article. I don't see why we are foregrounding NIST criticism that has not been discussed elsewhere. Nor has the paper as far as I can see been peer-reviewed. 210.7.71.114 (talk) —Preceding undated comment added 09:05, 14 December 2011 (UTC).

The section on Reliability of Biometric Methods is biased and against published academic information on biometrics. Specifically NIST section is irrelevant and unrelated to UIDAI. Criticism on the UIDAI's report is by the author himself and has not cited academic or biometric expert for his analysis. "Fundamental issues in..." paper cited refers to performance testing. It is an academic paper emphasizing opportunity to conduct more comprehensive measurement testing. It does not make any statement about reliability or usability of biometrics The last paragraph makes prediction of biometric industry, again unrelated to the UID project.

The section on Book on Aadhaar occupies far too much importance as it is one of numerous books, monograms and reports published on the project.

The criticism by the Standing Committee is factual and could be kept. — Preceding unsigned comment added by Rmashruwala (talk • contribs) 07:39, 20 February 2012 (UTC)

Last time I read the article was around June last year. It wasn't in a very good state then. And now, I agree with 210.7.71.114, it is desperately in need of re-writing. Brother Big (talk) 18:20, 28 February 2012 (UTC)

210.7.71.114 suggests that the NIST paper hasn't been peer-reviewed. That may or may not be correct. The recently-released UIDAI reports here and here also show no sign of peer review. Were the figures in those reports audited? What protocol was used for the trial? How does the reliability of NPR compare with Aadhaar? Brother Big (talk) 18:20, 28 February 2012 (UTC) NPR does not do biometric de-duplication. It merely is collecting biometric data.

It is correct that UIDAI reports are not peer reviewed. But they have been presented and are being presented in internationally recognized biometric conferences. — Preceding unsigned comment added by Rmashruwala (talk • contribs) 10:06, 2 March 2012 (UTC)

Thank you for that correction about the NPR. The census website confirms that deduplication is done for them by UIDAI, they don't do it themselves. Brother Big (talk) 08:52, 5 March 2012 (UTC)

Can we assume that the Rmashruwala commenting on this page is the Raj Mashruwala who wrote India boldly takes biometrics where no country has gone before? Brother Big (talk) 08:52, 5 March 2012 (UTC)

There is a clear case of Conflict of Interest - please see the resume of Raj Mashruwala quotes : "Chief Biometric Coordinator (2009-2011), UID Authority of India, Technology Centre. Mr. Mashruwala has been an early volunteer in this ambitious nation-building project of Indian Government. UIDAI plans to issue unique ID to every Indian resident based on comprehensive biometric trait." AshLin (talk) 14:04, 1 May 2013 (UTC)

Article history

For the record, Aadhaar was redirected here on 2 June 2011 by User:Bwilkins; last diff is here. Multipurpose National Identity Card was redirected to Aadhaar on 28 June 2011‎ by an anon editor and subsequently here. It wasn't great, but there may be bits worth recovering from the last diff.FlagSteward (talk) 09:33, 27 April 2012 (UTC)

CONSENSUS building

I have improved and trimmed the page to restrict text to UIDAI organisation and UID scheme. The page may still have biases. I have requested the other editor on his talk page to respect policies and from now on to only insert text for which there is consensus. Please use this section and make brief points to achieve consensus.Notabede (talk) 03:56, 23 October 2013 (UTC)

Copyright Vio

It has been brought to my notice for the first time today by "Admin:Spaceman_Spiff" that I (and perhaps other editors ) have apparently violated WP:C for this article. Although I believe that use of a small amount of well cited relevant text (citations as per the APA norms) from reliable secondary sources clearly falls under "fair use" WP:FU, nonetheless I am prepared to rewrite specific examples of text or place it within quotation marks to avoid any unwarranted imputations of being a habitual copyright violator etc. as and when they are posted on this page.Notabede (talk) 12:21, 23 October 2013 (UTC)

No copyright violation from my side

There is no copyright violation from my side. Please bring anything specific with proof to my knowledge. My writings are in my own words (no original research).

Infact someyhing Citing article ending with "NIRADHAAR" is not at all Cited by me and are trying to malign UIDAI. It is inserted by someone who protest against me.

Ravishyam Bangalore (talk) 12:16, 24 October 2013 (UTC)

Not edit warring

I had just deleted a claim which seemed to be essentially self published. I then noticed a revert sequence between 2 other editors on the same text. My last revert was not edit warring, just a routine edit. I'm stopping edits on this article for the next 48 hours to avoid any misunderstandings. Turnitinpro (talk) 15:10, 14 December 2013 (UTC)

Jan-2014

Deleted false and defamatory part. No Sale Deed and money payment done. Then how scam can happen? Government allocates its land and building for use to its various departments as a routine.

It is just a defamation.

Ravishyam Bangalore (talk) 04:22, 13 January 2014 (UTC)

Legality: RBI and Passport office cannot accept AN without legal backing. Bank accounts cannot be opened, Passposrt cannot be issued without legal backing. Hence deleted.

Some people want to block all positive information on the anti-corruption tool like Aadhaar from spread. They want only false, negative and disinformation. because they are afraid of anti-corruption drive in India which had been blocking its progress for centuries.

Ravishyam Bangalore (talk) 04:40, 13 January 2014 (UTC)