WiserEarth Editors

This is a wonderful feature in WE platform which is largely ignored by users, and this article is about bringing attention to it, explore the ways to make it more visible and therefore used, and how important it can become. I had figured it quite a while ago for user-to-user relationships, but only figured yesterday how generic it was and how interesting it could be in particular in the framework on current discussions on the various roles of users vs organizations.
When a user adds an item to its network, be it an organization, another user, an event, a resource, a wikipage ... an object called "relationship" is created in the data base. This object can afterwards be edited by the user, and two kinds of information can be added:
The nature/type/quality of the relationship, either picked in a drop-down menu, or specified as free text.

A short description of the relationship in a field "note". I've started to systematically qualify and document the relationships in my network. See
http://www.wiserearth.org/relation/view/bvatant/user/child_type/organization

For example one can specify a relationship to another user as "coworker", "friend", "collaborator", or a relationship to an organization as "member", "funder", "staff", to a resource as "author", "creator" etc.
This feature is extremely interesting. Adding this information is transforming simple links, like hypertext links, in qualified links. Links with added semantics. In the perspective of a publication of WE data to the Semantic Web, such information would be very useful. The ontology of relationships could be published, and the data base queried using those relationships, like "What are the resources ?R of which "author" ?A or "publisher" ?P is a "staff member" of organization X".
Why is this feature not used more?
First it's not trivial that it exists. Qualifying the relationship is not systematically proposed to the user when the relationship is created. The user has to list her network, figure what is behind the "Edit Relationship" action.
Second in the current interface this information is not leveraged in the search. One cannot filter the relationships by type. For example one cannot see only the resources of which a given user is author or publisher, or organizations of which the user is a staff member.
So, this very important feature is completely overlooked by most users. It should not be technically difficult to act on the two points above, and as soon as this feature is more visible and usable, communicate to users how important it is to use it.

 

Hi Heatherlynn

 

Thanks for jumping in.

 

Regarding veracity of self-reporting / free-editing. This question is inherent to everything published by anyone in any form of media. You are a librarian, you know those issues quite well. But I agree that maybe more than anything else, asserted semantics have to be tracked for source, trust, etc.

Putting semantics on people's relationships also raises issues of privacy. I discovered that you can edit a user's relationship "from outside", that is, from an object in her network. I was able to edit the relationship of Paul to Blessed Unrest, stating he is the author, which is not infringing much his privacy I guess, but you see what could be done using this feature.See http://www.wiserearth.org/relation/view/f8bd3e40733ff9c8e649298f53e49449/resource

So veracity is not only about self-description.

 

On the other point I agree about setting a well-thought but controlled set of relationship types, and avoid the free-tagging mess. OTOH users should be able to propose new candidate types to the editorial team, which could be added to the list through a transparent editorial process, similar to the way "candidate descriptors" are managed in a good thesaurus workflow. The current lists are clearly suboptimal.

 

Another, more technical issue is the direction and naming of relationships. Is you name "Authored by" the relationship between a book and a person, as in the above quoted case, it's quite clear who is the author of what.

But if you set "Affiliate" on a relationship between two orgs, it is unclear for a third party which one is affiliate of which one. The relationship should either indicate its direction, or be named differently from either side. In the current state of things, it's confusing.

See example at http://www.wiserearth.org/relation/view/67cf0d5bbf53b0c53a31040156f78c7c/organization

 

 

I admit I had no idea that that feature was even available as I have been madly testing and
discovering other parts of the system and trying to get a better grasp on where WE is in terms of
development.  I assume that this is also be true for others.

 

Camilla has it right, the personal profile pages are a bit cumbersome to make best use of this feature- in addition to tabs, it looks like the volume of info is pushing the need for collapsible and expandable areas. I will flag this thought for when I see the redesign issue come up, and I will watch for tHe cascading updates of tutorials, FAQs that relate to using this feature.

I am definitely only following Bernard's depth of knowledge on the semantics issue, and I look
forward to developments on this from a fairly critical "user's perspective" as well as from an
excited info-nut's perspective. I certainly agree that the editorial process be transparent and
widely inclusive of input with a clear process - WE seems to be developing this type of process
well, from my understanding. I also admit that what Bernard has pointed out about editing other's
relationships has me VERY interested in what people think about the likelihood of abusing this
particular editing freedom. Probably about the same as any other freely-editable item?

So other than placement, why would this feature go unused? User's may be drawn to this feature if
its use is made more apparent to their purpose and the manner of "controlling" the relationships is clear. Perhaps that is obvious, but I have been thinking about the return value of participating in WE from a recalcitrant view, and I keep ending up with the need to systemically MAKE users enter more detail when they begin, instead of just offering the ability to do so as an afterthought. I wonder if I am being a bit draconian, but I think the system will be better and more utilized if this feature is leveraged. Like I noted, I did not even know that this was doable when I first landed in the network.

My first thought about the *wiserearthling* use of the feature would be to expand the group messenging feature to select using the relationship.

thanks for teaching me as we go along!
~h

Hey guys, glad Bowo pointed out this thread to me - I'll chip in my thoughts and update everyone on any work that we do that is relevant to the topic at hand.  I worked with Angus in drawing up the current set of network relationships, and Angus had the vision initially to add semantic meaning to the network connections we had on hand.  The current network is still a work in progress - for our next iteration, we're looking to make the relationships be described in sentence form, rather than just a simple drop-down category.  I do wholeheartedly agree this is one of the lesser known features in this site and could become quite powerful with the right amount of nuturing and guidance.  We're meanwhile working our best to make this powerful feature become much more intuitive, and easier to set up.  Automated connections that are intelligently chosen based on the context in which it was made (for example, connecting someone to a resource as the record creator), and so forth.  I'm really glad to see there's interest in the community to explore how this could be used better, and I'll be on hand to continuously chime in my thoughts and follow up on the ideas laid out there!

At some point, I will begin drafting up some new proposals for how we will go about making this network feature more prominent and more easy to use, and also, more powerful.  I will say this - we have a pretty full pipeline of projects for 2009, starting with a site redesign, internationalization, and group refinements.  They may not sound like much, but there will be plenty of back-end design and extra elements added in that will string out each project.  So, I hate to throw out such a long timeline before we get around to networks, but I'll make very sure this stays atop WiserEarth's radar, and that it will be given its due weight in deciding what will be our next wave of projects!