Skip to content

Commit

Permalink
Merge pull request #1219 from edpars0ns/patch-4
Browse files Browse the repository at this point in the history 
Update index.html
  • Loading branch information
@jabhay
jabhay committed Nov 28, 2020
2 parents 4291e57 + bf51475 commit 0d69033
Showing 1 changed file with 78 additions and 76 deletions.
154 changes: 78 additions & 76 deletions responsible-use/index.html
View file
Original file line number Diff line number Diff line change
Expand Up @@ -101,7 +101,7 @@ <h2>Introduction</h2>
<h2>Context</h2>

<p>
Geospatial data is like a fingerprint: every combination of space, time, and theme is unique. The interaction with geospatial data can lead to beneficial insights and services, but it can also compromise citizens' privacy. This, in turn, may make them vulnerable to governmental overreach, tracking, discrimination, unwanted advertisement, and so forth. Hence, geospatial data ought to be handled with care. But what is careful, and what is careless? Let's discuss this.
Geospatial data may be seen as a fingerprint: For an individual every combination of their location in space, time, and theme is unique. The collection and sharing of indivudals geospatial data can lead to beneficial insights and services, but it can also compromise citizens' privacy. This, in turn, may make them vulnerable to governmental overreach, tracking, discrimination, unwanted advertisement, and so forth. Hence, geospatial data must to be handled with due care. But what is careful, and what is careless? Let's discuss this.
</p>
<p>
Even though W3C is famous for developing standards, this note is not normative. Rather than presenting an overview of best practices ready for implementation, the purpose of this note is to start the conversation. You are therefore encouraged to join the authors of this document in exploring the question: “How do we use geospatial data responsibly?”
Expand Down Expand Up @@ -131,6 +131,83 @@ <h2>Data Ethics</h2>

</p>
</section>
<section id='spatial-is-special'>
<h2>Spatial is Special</h2>
<p>
All domains are special. After all, this is what sets them apart from their neighboring disciplines. Space and
time, however, are special in how they cut across domains. They are the glue that helps us organize knowledge
about the world around us. We may be interested in events, such as landslides, that appeared within a region to
mitigate future risks, or we may be interested in events that happen during the same time but in a different
region, such as country-specific social distancing measures to slow the spread of COVID-19. In the first case,
space remains invariant; in the second case, time is invariant while we are interested in spatial differences.
</p>
<p>
Spatial data is special in many regards. The regional variability in the examples above illustrates two
competing properties of spatial phenomena. Spatial heterogeneity as a first-order property implies that it is
challenging to find observations that are prototypical for a particular region, e.g., which county is most
representative for COVID-19 transmission patterns. While spatial dependence as a second-order property between
observations violates traditional statistical theory. Samples are not independent; they covary as a function of
their distance. For instance, nearby sensors are likely to observe similar amounts of rainfall.
</p>
These two competing properties simultaneously drive the abilities and risks associated with spatial analytics, e.g.,
with regards to human movement. We assume that people frequenting the same places (or place types) may have other
similarities, such as common interests or demographic characteristics. However, great care has to be taken when
trying to assign properties averaged over a region to individuals, a problem known as the ecological fallacy. This
is particularly critical as modern information retrieval and recommender system utilize a person's location history
as a predictive feature. Frequently visiting a health care facility is not only indicative of health issues but also
of being a health care worker and those two cases may be difficult to distinguish.
<p>
Location information also increasingly triggers actions in the physical world around us , e.g., via geo-fencing,
or determines which information we receive, creating local information bubbles known as hyperlocal media.
Unfortunately, masking or perturbing location information comes with its own associated risks. For example, it
may place a user into a geographic area associated with another person. Similarly, users that share their
location, e.g., bike trajectory, may be mistaken for another person of the substantially larger population of
people that do not share their location in real time such as in a recent case involving a racially motivated
attack by a biker. Finally, the value of spatial and temporal information does not scale linearly with size,
once a certain spatial and temporal resolution has been achieved it becomes possible to monitor people and infer
their future behavior on a very fine grained level. This includes positive aspects such as finding missing
people or preventing crime, but also highly problematic cases such as targeting individuals, e.g., for political
or cultural reasons.
</p>
<section id='nature-of-spatio-temporal-data'>
<h2>The Nature of Spatio-Temporal Data</h2>
<p>
Geographical information, infomrtion about the physical nature of the world and about the social and econmic activities that take place may be described as Spatio-Temporal data, in that it relects the nature of the world at particualr locations at particular times.. In simple terms an atlas of Europe published in 1912 would represent now historic national boundaries, place names etc, in the same way a smartphone map of traffic conditions for San Francisco represents the amount of congrstion on roads a few seconds previously.
<p>
With the advent of smartphones and ability to collect and share infomrtion about indivuduals locations has become mainstream, it is the ability to place someone at a specific location at a specifc time that illustrate bith the value and the potential risks of such infomrtion. It is useful to think of time an space and two sides of a coin, and indivudal can be at only one location at specific time, and vice versa at a specific location at a particular time. In general terms if this data is abstracted or reduced in resolution both in terms of time and defintion of location the infomrtion value is decreased along with the inherent risk of exposing an indivudals location.
</p>
</section>
<section id='use-cases'>
<h2>Use Cases</h2>
<section id='use-case-1'>
<h2>Use Case 1</h2>
<p>
TBD
</p>
<ul class='note' title='Benefits'>
<li>TBD</li>
</ul>
<aside class='example' title='Example 1'>
TBD
</aside>
</section>
</section>
<section id='misuse-cases'>
<h2>Misuse Cases</h2>
<section id='misuse-case-1'>
<h2>Misuse Case 1</h2>
<p>
TBD
</p>
<ul class='note' title='Benefits'>
<li>TBD</li>
</ul>
<aside class='example' title='Example 1'>
TBD
</aside>
</section>
</section>
</section>
<section id='existing-legal-frameworks'>
<h2>Existing Legal Frameworks</h2>
<section id='ethics-and-law'>
Expand Down Expand Up @@ -269,81 +346,6 @@ <h2>Existing Ethical Frameworks</h2>
*A noteworthy exception to this are the Unicef ethical considerations. In order to protect children in unsafe regions, the use of spatial data is both essential and risky. Therefore a critical and detailed assessment of data is crucial.

</section>
<section id='spatial-is-special'>
<h2>Spatial is Special</h2>
<p>
All domains are special. After all, this is what sets them apart from their neighboring disciplines. Space and
time, however, are special in how they cut across domains. They are the glue that helps us organize knowledge
about the world around us. We may be interested in events, such as landslides, that appeared within a region to
mitigate future risks, or we may be interested in events that happen during the same time but in a different
region, such as country-specific social distancing measures to slow the spread of COVID-19. In the first case,
space remains invariant; in the second case, time is invariant while we are interested in spatial differences.
</p>
<p>
Spatial data is special in many regards. The regional variability in the examples above illustrates two
competing properties of spatial phenomena. Spatial heterogeneity as a first-order property implies that it is
challenging to find observations that are prototypical for a particular region, e.g., which county is most
representative for COVID-19 transmission patterns. While spatial dependence as a second-order property between
observations violates traditional statistical theory. Samples are not independent; they covary as a function of
their distance. For instance, nearby sensors are likely to observe similar amounts of rainfall.
</p>
These two competing properties simultaneously drive the abilities and risks associated with spatial analytics, e.g.,
with regards to human movement. We assume that people frequenting the same places (or place types) may have other
similarities, such as common interests or demographic characteristics. However, great care has to be taken when
trying to assign properties averaged over a region to individuals, a problem known as the ecological fallacy. This
is particularly critical as modern information retrieval and recommender system utilize a person's location history
as a predictive feature. Frequently visiting a health care facility is not only indicative of health issues but also
of being a health care worker and those two cases may be difficult to distinguish.
<p>
Location information also increasingly triggers actions in the physical world around us , e.g., via geo-fencing,
or determines which information we receive, creating local information bubbles known as hyperlocal media.
Unfortunately, masking or perturbing location information comes with its own associated risks. For example, it
may place a user into a geographic area associated with another person. Similarly, users that share their
location, e.g., bike trajectory, may be mistaken for another person of the substantially larger population of
people that do not share their location in real time such as in a recent case involving a racially motivated
attack by a biker. Finally, the value of spatial and temporal information does not scale linearly with size,
once a certain spatial and temporal resolution has been achieved it becomes possible to monitor people and infer
their future behavior on a very fine grained level. This includes positive aspects such as finding missing
people or preventing crime, but also highly problematic cases such as targeting individuals, e.g., for political
or cultural reasons.
</p>
<section id='nature-of-spatio-temporal-data'>
<h2>The Nature of Spatio-Temporal Data</h2>
<p>
TBD
</p>
</section>
<section id='use-cases'>
<h2>Use Cases</h2>
<section id='use-case-1'>
<h2>Use Case 1</h2>
<p>
TBD
</p>
<ul class='note' title='Benefits'>
<li>TBD</li>
</ul>
<aside class='example' title='Example 1'>
TBD
</aside>
</section>
</section>
<section id='misuse-cases'>
<h2>Misuse Cases</h2>
<section id='misuse-case-1'>
<h2>Misuse Case 1</h2>
<p>
TBD
</p>
<ul class='note' title='Benefits'>
<li>TBD</li>
</ul>
<aside class='example' title='Example 1'>
TBD
</aside>
</section>
</section>
</section>
<section id='developer-perspective'>
<h2>Principles of Ethical Data Sharing - The Developer Perspective</h2>
<p>
Expand Down

0 comments on commit 0d69033

Please sign in to comment.