Wow! This is great.
Regarding open data definitions OKNF 10 and the 8 principles miss many of parameters such as the following which are normalized practices in the natural and social sciences:
- Data quality (accuracy, reliability, integrity, Identity, authenticity, completeness, semantics, sound methodology, etc.)
- Description (metadata - provenance, dates, versions, forms, attributes, collection methodologies, quality parameters, classification explanations, authors, and so on) - this helps with fit for use decisions, provides trust in the data, and enables discoverability.
- Interoperability (legal, technical, semantic, policy) - where laws, technologies, classifications and policies enable use and reuse
- Standards (metadata, dissemination, services, interoperability, etc.)
- Preservation (archiving, systems to archive), we have the data today but we also want them tomorrow
With regards to open data and context. It is true that context matters in terms of the release of data, for example, health data are not released because of privacy issues, census data are aggregated for the same reason, broadcasting the location of endangered species may lead to their harvesting, mapping bear dens may invite unwanted tourists, sharing sacred aboriginal sites may lead to them being de-consecrated, archeological sites looted and so on. Ted, I think, one of the things you are grappling with is the decision making process on how to release sensitive data.
GeoConnections has produced a report to address issues pertaining to the release of sensitive data. The sensitive data are defined as follows:
- Legislation/Policies/Permits – the data is identified by legislation as requiring safeguarding. The most prominent legislation in this regard is the federal Privacy Act – safeguarding the data is required if an individual can be identified, either directly by georeferenced information (such as the geo-coordinates of an address) or indirectly through the amalgamation of geospatial data and related attributes;
- Confidentiality – the data is considered confidential by an organization or its use can be economically detrimental to a commercial interest;
- Natural Resource Protection – the use of the information can result in the degradation of an environmentally significant site or resource;
- Cultural Protection – the use of the information can result in the degradation of an culturally significant site or resource; or
- Safety and Security – the information can be used to endanger public health and safety.
The principles for the release of sensitive data are as follows:
The info above is from the Best practices for sharing sensitive environmental geospatial data which on page 27, provides an excellent Sensitivity Assessment Procedure decision making tree that I think all of us interested in open data should take into consideration, and data creators especially, as the decisions on what to release and what not to release to the public needs to be more transparent and objective. I think a more general decision making tree needs to be developed that will also include contractual obligations and IP would be most helpful. I hope you will take the time to read it. Other excellent operational policy primers related to data are available here - http://geoconnections.nrcan.gc.ca/18.
- Unless the dataset is classified as sensitive it can be provided free of restrictions;
- Information can not be considered sensitive if it is readily available through other sources or if it is not unique;
- The Data Custodian of the information is the only agency that can determine whether an environmental geospatial dataset is to be classified as sensitive;
- Data Consumers of sensitive environmental geospatial datasets must honour the restrictions accompanying the information in the form of an agreement, license and/or metadata; and
- Organizations should document and openly publish their process, criteria and decisions.
Heather, CIPPIC has mapped open access licenses in this report - http://www.cippic.ca/sites/default/files/Open_License_Comparison_Report-v2-10Feb2011.pdf, and here - http://www.cippic.ca/sites/default/files/CIPPIC-Ottawa_License_Report-2010-11-15.pdf. This was work I inspired some time ago when there were only 4 open data cities, and I had significant concerns about the proliferation of not so open licences. Some discussions related to these can be found here https://traceyplauriault.wordpress.com/2010/07/21/changecamp-ottawa-2010-open-data-terms-of-use-session/. Might be good to check with Kent and David at CIPPIC to ensure you are not reinventing any wheels.
The next item Ted, that I think you are addressing is the clash between local and occupancy, land ownership, treaty rights and unfair/uneven exploration and exploitation combined with the maverick like tendencies of surveyors and geologists. These are issues beyond the scope of open data.
However, where open data and these issues do intersect is local and traditional knowledge (LTK) and IP, and western legal frameworks of IP protecting individual rights but not collective rights. The Lab where I work is doing some work here, with CIPPIC and Teressa Scassa (https://gcrc.carleton.ca/confluence/display/GCRCWEB/Mapping+the+Legal+and+Policy+Boundaries+of+Digital+Cartography), and we are in the process of developing a LTK licence that takes into consideration OCAP principals (ownership, control, access, and possession) of aboriginal information (http://www.rhs-ers.ca/node/2), combined with the sensitive info issues discussed above, and based on a number findings over the years in our mapping work in the north. We have heard loud and clear from Inuit, that they want control over their information, their local and traditional knowledge, and we are trying to find ways to incorporate those concerns into a creative commons like license. In addition we hope to build capacity in these areas and ensure that government and the private sector follow these principles, adhere to the licences and that aboriginal people know and are able to enforce their rights.
Sorry for being long winded. Your post got a conversation going and that is a good thing and I hope some of what is in this response proves helpful.
On Tue, Feb 26, 2013 at 4:46 PM, Ted Strauss <[hidden email]> wrote:
I want to concede that "to achieve one's goals in the context of a particular situation" (my words)is a completely subjective statement, and so ineligible to be part of any formal definition.I acknowledge that my post emphasized definitions more than was warranted.This discussion has helped clarify for me that one reason for my post was a sensethat the close bond between the license/definition writers and the movement itself has led toless discussion of when to open or not open data. Take this quote from http://okfn.org/opendata/under the heading "How to open data" (bold in original):
"Address common fears and misunderstandings. This is especially important if you are working with or within large institutions such as government. When opening up data you will encounter plenty of questions and fears. It is important to (a) identify the most important ones and (b) address them at as early a stage as possible."Compare that to the following from http://opendefinition.org/about/ (both quotes by the same organization)"The definition sets forth principles by which to judge whether a knowledge license is open, it does not seek to provide or recommend specific licenses."So while the definition writers do not recommend specific licenses, on another page they also want to remind their advocates how to address common fears and misunderstandings about the same licenses. I think OKF has gone to sufficient lengths to separate these two sections of their site so they reach the appropriate audience; I am not crying hypocrisy. But with direct advocacy on one side, and objective licenses on the other, where does substantive discussion fit in about when is it recommended to apply these licenses? I haven't heard enough of that discussion and if nothing else, I think cases like the Weenusk First Nation should encourage more of that discussion.Maybe there should be an openmotivation.org with a decision tree helping governments and organizations decide how to make use of open data (if at all).This could help them decide what goals are achievable, and what contexts can help direct their efforts most effectively.I will make an example of what I mean by this and post it soon.Thanks again everyone who read the blog and posted comments.TedOn Tue, Feb 26, 2013 at 3:27 PM, Michael Mulley <[hidden email]> wrote:> Given that there is an underlying bias that open is desirable,> shouldn't there be a discussion about what conditions make> openness desirable or not.
Oh, for sure!
We've had a bunch of discussions here before about whether specific examples of released data were good or bad, and these are usually very interesting. An attempt to formulate a broader list of conditions that might make openness undesirable could likewise be interesting.My & others' objections, I think, were to muddying definitional waters, and to the Weenusk example, which seems to me not a question of whether data should be open but whether it should've been collected in the first place (and then, if one answers no to this question, whether it should be kept private -- its licensing doesn't seem like the relevant question here)._______________________________________________
CivicAccess-discuss mailing list
[hidden email]
http://lists.pwd.ca/mailman/listinfo/civicaccess-discuss
_______________________________________________
CivicAccess-discuss mailing list
[hidden email]
http://lists.pwd.ca/mailman/listinfo/civicaccess-discuss
--Tracey P. LauriaultPost Doctoral FellowGeomatics and Cartographic Research Centre
_______________________________________________
CivicAccess-discuss mailing list
[hidden email]
http://lists.pwd.ca/mailman/listinfo/civicaccess-discuss
| Free forum by Nabble | Edit this page |