Open Data models/collaboration, practical discourse, practical work

[Michael Maranda]

I'll be speakng to the choir here, I know.

I posted a comment on the CatComm page at WiserEarth, in hopes of cross pollinating some ideas. http://www.wiserearth.org/organization/view/fd162fbfff4c9d6c9c7b005...

I'll repost what I wrote there, perhaps with minor changes. I'm hoping to do a longer piece or series of blog posts on this topic, but I am more interested in the work progressing than in qwaxing eloguent (or obtusely, as the case may be). I'd rather the dialogue started and that we got to the business of making our data models public, inviting others to take them up and use them and to offer feedback that allows their refinement. This is what I am committed to under the aegis of a thriving OSN.

How does one most effectively find peer organizations on WE? What if the peerage is on a meta-level? CatComm.org is one among many pursuing similarly broad aims to WE. We're interested in establishing common public standards for the field and we're committed to the exchange of data and the evolution of an ecology of solutions and cooperative data management.



We've looked to WE as an exemplar in creating the directory or index of organizations in the movement with no name, as we've tried to provide a place for the documentation of community solutions (at CatComm).



Other groups are working on similar efforts, and we recognize that we're better off working together to define specific use cases where data cooperation can be meaningful rather than abstractly calling for open APIs.



Our new platform, (soon to launch) will allow access to all our content via a ReSTful API.



We propose that what is needed is a public space for dialogue on mutual efforts such as the Open Sustainability Network. We begin the process with invitation and we ask that others, such as the WE leadership and community, join us in issuing the invitation:



Where best to begin? Take the data models already in use, as (for example), the data model or format of the WE index... explicitly declare it a public data model, explaining the choices made and inviting constructive criticism so that it may be further refined. Also, clearly state the model to be intended for common use that other initiatives may freely use or build upon.



The aim is simple: in order for the API to be meaningful, we align ourselves to a public data model. The evolution of the model must be safeguarded - under appropriate stewardship.



On this basis we begin to share data throughout the network, with multiple points of data-entry and data-access.



We are speaking not of one data model, but of several. Each logical data grouping is a data model in these terms. The data points on organizations would constitute one such model. Data points on solutions represents another modei. Let's build a netowkr ecology together.

[Chris Watkins]

Good, good.

Have you made contact with the WE folks yet? Lonny and Curt talked with them a while back, and the general sentiment was that we wanted to cooperate, but we need to follow up.

WiserEarth's content is under a CC-BY-NC-SA license - too restrictive for Appropedia to use. HOWEVER I don't know if that applies to basic data in lists - I understand that under US law, the information in a phone book cannot be copyrighted, as it's not a creative work.

I would like to play with WiserEarth's data, and do things with it that I can't do on the WE site itself. I would like to know:
* Can we access WE's data, in some kind of dump?
* If so, do we have any license issues if we used any part of it on Appropedia or CatComm?
* If not, where did WiserEarth get their initial information, and can we access this and use it (making it open, of course)?
* How do we safeguard the numbers and emails of people that we might find in this data? WiserEarth has the emails on their website displayed in a slightly non-raw form - I'm not sure how secure that is from scraping by spammers' bots, but I'd also be concerned about exactly how emails are displayed in any output we provide as part of being open.

[Michael Maranda]

I spoke with Angus Parker several months back (prior to our OSN gathering - iin follow up I was hoping others leading OSN effort would also touch base with him). My discussion with Angus was very positive. I basically suggested steps along the lines described in the post here, and I thought we had basic agreement on these points.

There is always the idea of an open API ... and I hear they are still working on that for WE (and perhaps seeking funding for same).

My bigger point is that it's important to take other steps first when you are trying to advance a field.

My general sentiment is that an "open API" is an abstract good, but establishing clear invitations at collaboration and clear invitation to share, and practical working exchanges is much more important and will lead to actual use rather than a nice notion (and potentially uncertain implementation or uptake).

What I took away from the conversation was that the data model for Wiser Earth (namely on the particular set of fields utilized in the "index" at the core of the WE project) could be considered a public standard and that others could go ahead ad use the same set of categories.

I asked that they find a space on WE to make that a clearly stated invitation, and I still think that is the most important step they could take towards establishing a collaborative space between our several endeavors.

From alignment on basic data models we become more useful to each other.

[Chris Watkins]

I did try to call Angus after talking to you about this - but have been busy, and I wasn't persistent with it. I still haven't emailed.

[Michael Maranda]

Chris raises a number of important and particular questions that I dont have the answer to, not being privy to the WE's internal discussions...

I hope folks from WE are joining us on Global Swadeshi.

As for where the WE data itself came from - I expect the answer is multiple sources. I understand they had staffing resources for the collection of the initial data.

I understand they are either working ion an API or on funding for an API. But that's a very general concept. The specifics are important. What data will be available, and under what terms? There are many communities of practice that would benefit from such openness.

I do understand the points re: email and protecting the email account holders from spam/harvesting. I dont know their policy or what means they are taking to address that point.

Which parts of CC-BY-NC-SA are difficult for Appropedia? the NC part?

[Chris Watkins]

Yes - NC, the non-commercial clause, is the problem. Not just that we don't agree with it, but it's also not compatible.

We could mix the use of the licenses if the actual content doesn't mix, but we couldn't take text from their site and use it in Appropedia, and vice-versa. (Lists of data, such as the content of phone books, can't be copyrighted in the USA, so some of the basic info should be fine to use in any way.)

[Michael Maranda]

This raises an important point - in some cases granularity of content terms would appear appropriate. The bulk of user generated content - in threads rather than in documented solutions for example might be best serve under NC terms. Solution models, and other "public data" - possibly the NC license would not apply. Deciding what policy for what element or category of content requires some justification.

[Fabio Barone]

I am not very familiar to the discussion of unifying data models, so please allow me some considerations:

- You mention a "network ecology". I understand a network to connect nodes on a common communication protocol. What is behind the node is implementation specific and therefore "private" to the node. In other words, Internet works because of the TCP/IP protocol; a node may be Windows, Linux, Mac, a Pentium II or a supercomputer, etc. This approach has proved to be very successful. I do not immediately see why you want to start from the data model?
- I understand your reserve in calling for open API, as this could imply an elaborate process.
- Still, I would rather see to model the interaction (functions, REST) first and leave the data model for later. Even with a unified data model you need to define what to do with it. Who is going to use the functions? What is going to be done with them? (User stories, use cases, you name it).
- What makes WE's data model the choice for your ideas? I had heard of WE before, but don't know it really (maybe a USA - Europe thing). Why should the data model be taken from there?

I have thought several times of connecting activists,alternative,cultural creatives, etc. together. Where I stumbled was always about "how does it work"? WHAT information is really the one you want to share? How do you do it? How to ensure there is no overload of information?

Hope my questions are clearly formulated...

[Michael Maranda]

Several things to address to clarify what I am saying. I may only cover a few in this post (it's late here)

First, the sense of network here pertains to the network of people and organizations, and of other networks. One text I recommend is "Movement as Network" by Gideon Rosenblatt... but that's an aside.

My notion of data model here ... centers around an object-category or type. Think in terms of not of one but of multiple models for any complex effort - seeking models for each domain or type of object we'd like to share data about. ... i.e. this is not requiring total alignment between sites on all data they are working with, but we can group certain categories of data...

"Organization" is one such entity -- it's a core data model in the WE project - styled as the "World Index of Social & Environmentally Responsible Organizations" ... another tool or data model of the WE project is their category system for these organizations ... their "Areas of Focus" (AoF). WE is more than this data set, it's also a social-network, online community. The conceptual unity of all this is that the many Social & Env. Responsible entities world-wide constitute a movement, and that it's important to document this movement.

In addition to the Index of organizations - they're also providing means of tracking or documenting a number of other things of interest -- "Solutions" being one.

Going back to the phrase "network ecology" - my notion is that the network perspective is important in addressing strategic choices for the field of practice, for thinking in terms of a commons and in coordinating activity in a space where hierarchical control is neither desirable nor practicable.

There are many cases where this is true and where we operate ourselves as if it isnt. It's a challenge enough to reach the perspective of the network view. It's a further challenge to begin to conduct oneself from this view (and to keep organizational interests and conflicts present in mind). A lot of this blends with insights gained in the practice of open space and the F/LOSS community, not to mention analogies and lessons of the Internet in general. (Wiki has a place in this too!)

In the social-good sector (for want of a better term, tho I do like "movement with no name, or movement of movements, forgive me) getting ourselves to the wider, more encompassing frame and not allowing ourselves to emulate mindlessly the behaviors of other sectors where branding or stickiness or keeping information in proprietary formats or any number of other ideas we'd question in an open web environment - is not an easy task. The impulse to replicate old models is an obstacle that is hard to overcome.

Many sites that would ostensibly be in alignment on social aims find themselves in some degree of competition with each other, or at the least, duplicating efforts. For me, the OSN is an expression of a will to do better in these circumstances.

In the realm of limited resources ... in online terms ... in the information sector ... information ought to more easily flow, and our human attention should be considered precious.

When someone visits one of the many "social good" sites (WE or other sites, there are many) two things that may bring them there (not an exhaustive accounting) are the desire to a) document something they are doing or b) search/find information others have documented.

The problem we face in the current situation is that persons or non-profit and other organizations are compelled to either make a choice as to where they'll put their energy in representing themselves on these sites and whatever social networks, or similarly, when searching for such information... how many sites they'll search.

This example of the basic problem is focused on "organizations" .... but the general question can be applied to the core organizing content of Appropedia (Appropriate Technology Models) or CatComm.org (Community Driven Solutions) or any number of other related topics.

Speaking for CatComm... while we serve as a place for communities to self-document "solutions" they have attempted ... we know other groups have come up with the same basic idea and have developed parallel projects. Additionally - tho our core organizing idea is the "community solution" it makes sense for us to make it easier and more or less standardized for groups to document or identify their "organization". But we don't want them to duplicate their efforts over numerous sites ... well, we dont mind the fact of their content being in multiple locations, more to the point, we'd love for them to be available in multiple places online without having to populate and police these multiple spaces. So, in the area of "organizational profiles" it would be great to have an equivalent to "Open Social" for organizations of the sort that WE is tracking.

Why do I point to the WE model? In part because of scope of the index/directory.

However my notion of a "Public/Published Data Model" is one where there is room for ongoing refinement of the model. So, I am not suggesting a WE represents a perfect model. Indeed, I am not pointing to WE as a total model, rather I am starting with two parts-- one -- what data points are to be gathered for organizations - and are we (in other groups/sites) tracking similar data points & is there value in bringing some alignment so that we can improve the wider data pool and - two - their category system - the "Areas of Focus" as a means of grouping these Organizations. There are additional data models in the WE project, to be sure, but these are the one's I have focused my attention upon - as I see distinct use cases / user stories - where we can cooperate and build together, or build according to a network ecology where we come to value each other as resources.

I have gone on a long time in this response, and surely have not addressed all your questions - perhaps I have also raised new ones, or gone down a few rabbit-holes. Happy to answer further or clarify, or to be called out for getting things wrong or neglecting items of concern.

[Michael Maranda]

Short note -- I am very much in alignment with the point re: functions/ReSTful architecture, user stories... that's my point in getting specific rather than speaking idealistically about an Open API ... getting to the nitty gritty of what is shared and why. I take the "Openness" part as necessary, desirable, and hopefully a given ... though I know not everyone has bought into that (even if most of us here have). What is most useful to share, and what is the mode of sharing that will be most easily leveraged? What decisions about data collection, sharing etc. will enhance the network rather than just one's position in a network?

[David Braden]

Michael said:

getting to the nitty gritty of what is shared and why.

The feature of Wiser Earth - and its greatest limitation - is the Areas of Focus. It is a great tool to find all those who think the way you think. I much prefer the CatComm vision of a community coming together to solve a common problem. (Wiser Earth does have a feature where you can find all the listed organizations in a given locality. That feature will be useful when people start to look up from their individual projects to see how what they want to do fits with what other people want to do.)

The problems we face are systemic and cannot be solved one at a time. So the first step is a conversation across interest and expertise - and I would focus that on local systems of production because, large complex systems, like life on earth, require stable component systems like Community Investment Enterprises.

When we start talking about the different ways people, plants and creatures interact in our locality, we will want to know who has the expertise on each interaction - and a way to bring them into the design process.

[Fabio Barone]

I apologize, guys, but I still do not understand what you really want to do - and especially how you would like this discussion forum to help...sorry...