ontolog-forum
[Top] [All Lists]

Re: [ontolog-forum] Ontological issues relative to privacy.

To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: Ali H <asaegyn@xxxxxxxxx>
Date: Thu, 16 Jan 2014 14:13:03 -0500
Message-id: <CADr70E0ij6TL_tgGK16wYi60_=XW4OYbjLZ1jUKPx_swx-Pj0g@xxxxxxxxxxxxxx>
Dear John,

I think it is possible to identify a core conception of privacy that is extended in various ways by different actors.


On Thu, Jan 16, 2014 at 12:15 PM, John Bottoms <john@xxxxxxxxxxxxxxxxxxxx> wrote:
All...

There are various views on what constitutes privacy. Ultimately it comes
down to ethical and legal definitions.

At its core, a simple definition of privacy might be as follows:

The expectation for an individual that the signals the individual generates (whether talking in a room or by explicitly recording signals on some medium) will only be accessible to intended entities.

Specifically, imagine that in some time interval, a given individual engages in some set of activities which generate some set of signals. For example, my writing this email at a computer generates a slew of signals across multiple spectra: The actual typing of my email message is one set of signals; but also the sounds of the tapping of my fingers on the keyboard; and also the light bouncing off of my body, any facial expressions I make, etc. 

One can then extend or justify the above according to myriad ethical, legal and democracy oriented appeals. For example, people who value human rights and perceive a connection between privacy and functioning democracies might argue that some level of privacy is a right bestowed upon living things (humans), and having spaces where you control access to signals you generate are vital to freedom of thought, _expression_ and exploration. Others might conceive of it as a commodity that individuals can monetize or engage in commercial transactions with. Obviously there are many interpretations and extensions given those set of considerations.

At a technical level though, it's worth pointing out that we're at a unique period in humanity's evolution. The intersection of:
  • nearly ubiquitous internet and computing devices, 
  • greatly expanded computing power, 
  • poor (or purposely undermined) signal security protocols 
coupled with actors with large budgets and motivation to have "total information awareness" (and especially coming from a SIGINT community background), means that there are entities who are capable collecting and analysing a significant portion of all human generated signals. 

This is certainly unprecedented in human history, and a very fundamental conception of privacy that we've taken for granted (arguable since humanity has ever existed) is potentially undermined. Coupled to this of course is that privacy (especially in the computer age) is not a very tangible right, it's loss is not necessarily immediately experiential, and there has been a re-normalization process underway as a side-effect of certain business and economic models that have been that have been enabled by at least two of the technological factors listed above. (A brief aside - a friend dressed as "the NSA" for halloween this year and walked around with a microphone and a recorder at a halloween party. Without fail, every person in each conversation he entered objected to his recording of the conversation, where conversely many people seem to take it for granted and do not object that some/many/all of their conversations are indeed recorded by some entity who is certainlynot necessarily their friend) 

We can further discriminate between two types of privacy interactions:
  1. entities which you willingly give access to some set of your signals (generally through some agreement, though consent might be informal, say simply by engaging in correspondence / conversation),
  2. entities which access your signals without consent
In the first instance, it's further important to note that the balance of power and information between the two parties is rarely equal. Google can easily change its terms or service or privacy policy, and your only recourse is to stop using the service. They can unilaterally alter the terms of engagement, you cannot. They have a team of legal or other paid staff to construct policies aligned with their strategic interests, you probably have neither the legal background nor time to fully read let alone understand the implications of such agreements.

In this context, the role of ontologists in the first instance might include:
  • Help better balance the disproportionate power/information positions of entities.
  • Better articulate and making accessible elements of privacy policies.
  • Is there some classification of privacy terms that can be generalized across privacy contracts? 
  • Is there some way of making the effects of these accessible to end-users?
  • Is there a way to help end users evaluate the effects of changes in privacy policies?
In the latter instance, there are undoubtedly coupled political considerations, but some technological / computer-related opportunities might be to use ontologies in creating or enforcing more secure computing practices.
  • Ontologists might also help capture or make more tangible the benefits / costs of privacy. 
  • Ontologies can certainly be used to demonstrate how signals picked up by some entity can form incomplete pictures of surveilled individuals and might lead to erroneous or mislead inferences about them based on collected signals.
  • One can conceive of ontologies which list different types of signal generating activities and what surveillance techniques (might) exist for them and potential countermeasures. 
In both cases, as you note in bullet 1 below, there can be utility for ontologies in some degree of policy automation.

... I hope this is on topic.

Best,
Ali


In the meantime commercial
entities try to push the limits in a game of brinkmanship that coerce
individuals to sign their privacy away and then sell information derived
from that access it to their trading partners.

The prime profession in statistical assessment is the psychometrician,
one who verifies that a test performs as purported and does not
misrepresent the results of the test. Psychometricians put their
credentials on the line and have ethical guidelines because their
findings may be the basis for later work which may have legal consequence.

I would like to consider the following questions:

1. If there is a legal basis for privacy information requests and you
are asked to develop an ontology that implements only a portion of that
basis, what should your response be?

2. Is there an imperative for ontologists to develop a code of ethics to
communicate their role?

3. Should ontologists identify situations that may have legal
consequences and prefer to act in an advisory role to SME's rather than
actually drafting the ontology?

-John Bottoms
  FirstStar Systems
  Concord, MA USA


_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J




--
.
(•`'·.¸(`'·.¸(•)¸.·'´)¸.·'´•) .,.,

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J    (01)

<Prev in Thread] Current Thread [Next in Thread>