yes, I think we could find interesting results by agreeing to push a single shared tag in delicious -- I'll second 'nonprofit_tech' as a good term to try. I think self-organizing taxonomies can be too arbitrary; and that constructed taxonomies can be too artificial.

I would ask what is the purpose of the taxonomy? Is it to define, organize and describe the relationships between entities in a field? What is the value and what are the goals of this taxonomy? Will this create splinters and fracture lines and/or wrongly lump things together?

If you know of any bright library science masters who could help us wrangle these things, please invite them.

I like the idea of building our field and want what we do to support this.

Beth Kanter put together a list of shared bookmarks on nonprofit_tech, some time ago, and I posted it at http://mist146.drizzle.com/bookmarknpotools.htm -- food for delicious anyone?

Regarding the TechFinder taxonomy, the original idea was for a community or committee to work together to maintain the taxonomy; to add and revise to it actively over time. That committee/community hasn't (how shall I put it?) self-organized very much, so mostly this taxonomy has been maintained by TechSoup and N-TEN. Rem may have more to add on this. I can post the current TechFinder taxonomy here when I get a chance. Taxonomy creation always encodes a certain way of thinking; which makes me shy of taxonomies. All taxonomies are in a sense predjudiced, so you have to work against that in delicious and elsewhere. Nonetheless, I keep needing to find relationships between ideas and work, to put thoughts and things together, and to find ways to see parts and wholes and how they fit together, and taxonomies help with that.

The TechFinder taxonomy is a constructed one for a field that is new and doesn't have many established standard categories, and it's more useful after you use it for a while and learn it; IMO it's not immediately grasped in first use. It can conform to standards, which can be very good. It was designed to grow and deepen with additions (the taxonomy and application design supports growing more branches beneath any facet, though I don't think there's been much interest in growing this). It would be great to mine some good taxonomy entries that emerged from delicious and spread that around.

I just want to pipe in and say I appreciate what I am learning so far, in terms of context, etc.

One author that echoes your points is the Sociologist Pierre Bourdieu (a favorite of mine).

Is it fair to say that a Taxonomy implies a purpose (or set of purposes)? How broad can the purpose be, or how many if plurality of purposes can work?

If a group is working on a taxonomy for a website for an organization, I fear this is bound for difficulties if its an organizational tool for a diffuse organization rather than a means of organizing specific projects, or types of projects.

If the group brings together diverse, is the activity of generating a taxonomy a process of unifying their perspectives?

nonprofit_software is another taxonomy entry I'd propose we use. should we consider adding a nonprofit_* or nonprofit_tech_* prefix to anything we think deserves taxonomy level status? npo_tech or npo_software or np_tech would save characters. uh oh, are we going to need to make a decision?

I think phil is on to something here.

I propose that top-level tag is (to save characters):

nptech

Users are encouraged to innovate at that level below:

npotech_software npotech_opensource npotech_asps whatever...

Eventually we might get some agreement on a second level taxonomy.

With del.icio.us you wouldn't actually need the extenders. We could bundle anything that came w/ nptech.

So you could have:

nptech software fundraising

or

nptech consultant Michigan

We could sort on anything with the "nptech" and then look at the list of related links to see the other tags that people have added.

A bit of population and we'd probably end up with a good 2nd level taxonomy.

As for what we applying a taxonomy to -- why don't we let that emerge as well? What are people bookmarking? When I start to think of these things, I immediately jump to software but maybe people are putting other things there -- innovation or projects or articles or possibilities. Who knows?

If we go with nptech, I recommend that we push it out in all the usual places and then start figuring out how do the next step -- manipulating the data.

As kramer from Seinfeld once put it, it's all about levels. we could put everything in one uniquely defined bucket-- nptech, and then everything below that is in more natural language; or we could have additional uniquely defined terms ie: with prefixes. I think starting with nptech is fine. We can consider later if add'l terms are needed.

As a cheap experiment with potential, i think this idea is a good one. In the longer run, I think we should consider some broader issues:

1. Maybe np, the broader nonprofit sector, would be worth considering, rather than couching things in nptech. imo, there's already a pronounced tech bias in delicious.
2. limitations of delicious as a tool. What are these and how will these affect the success of this project? multi word taxonomy entries aren't possible in a consistent manner, the 'earlier/later' dichotomy in delicious may not be an appropriate filter on quality and relevance. what's the potential for gaming the system by a small number of people with uncommon interests?
3. audience. I think we should be clear about the intended audience. IMO, i think that the nptech taxonomy is intended primarily for use by people who work in this field, and secondly by np staff. Maybe this isn't right.
4. goals. a few goal suggestions:

• to have a well-organized, self-maintaining bookmark library that we can share.
• to discern a taxonomy by observing how nptech's classify resources
• to get a sense of the quantity & diversity of web resources within this nptech field.

OK, so go to http://del.icio.us/tag/nptech to see the taxonomy. We should suggest that people use the tag: nptech whenever they add relevant entries to delicious. To watch the taxonmy grow, subscribe to this site http://del.icio.us/tag/nptech as a delicious user.

What should get tagged as nptech? here's my suggestion: we use the same standards for inclusion as TechFinder. Here these would be suggested, unenforceable guidelines.

1. Sites must be directly relevant to nonprofits' information technology needs.
2. Sites must be technology-related.
3. Language used listings must be appropriate.

in more detail, snipped/paraphrased from http://techsoup.org/techfinder/index.cfm?p=about_standards :

1. Services must be directly relevant to nonprofits' information technology needs. Typically, this means that your services are either specifically designed for nonprofits, or discounted for nonprofits.

2. Services must be technology-related. NPTech is specifically designed to help nonprofits find support with their information and communication technology needs. Please ensure that your services meet this criteria.

3. Language used in NPTech listings must be appropriate. Do not use overly promotional language.

a question re: using delicious: what's the best way to use the 'extended' tags, and how do they function? Any general tips for using delicious as we start using it for nptech?

One tip is that you can rename taxonomy entries--so if you'd been using npo_tech or something else, you can rename all your tagged links to be nptech within delicious.

Here's an article that might help throw some more fuel on the fire of this discussion: Folksonomy - more collective classification.

Phil Klein said:

What should get tagged as nptech? here's my suggestion: we use the same standards for inclusion as TechFinder. Here these would be suggested, unenforceable guidelines.

1. Sites must be directly relevant to nonprofits' information technology needs.
2. Sites must be technology-related.
3. Language used listings must be appropriate.

in more detail, snipped/paraphrased from http://techsoup.org/techfinder/index.cfm?p=about_standards :

1. Services must be directly relevant to nonprofits' information technology needs. Typically, this means that your services are either specifically designed for nonprofits, or discounted for nonprofits.

2. Services must be technology-related. NPTech is specifically designed to help nonprofits find support with their information and communication technology needs. Please ensure that your services meet this criteria.

3. Language used in NPTech listings must be appropriate. Do not use overly promotional language.

I agree with the idea of making the data collected about the intersection of nonprofits and technology. However, I'm ambivalent about the idea of only pointing to entire websites.

One of the interesting things about del.icio.us is the ability to point to anything with a URL -- are case studies -- a single page a website useful -- how about an article about something that relates to nptech on the Washington Post? A picture of a Community Technology Center?

One of the benefits of opening this up could be, quite simply, that we begin to find out what kind of information people are collecting and the natural language ways in which they describe this.

I can easily imagine this data being used to jump start all kinds of projects: new navigational design for TechSoup, a way to create meaningful RSS feeds for various nonprofit users, words to put into meta tags of websites and web pages so that they show up on Google searches.

The key, I think, will be a concentrated analysis of the collected data and then publication of the results.

Thinking out loud here. What do others think?

Regarding http://del.icio.us

Aside from some avid users of this bookmarking site/tool, do we have any of the site developers/proprieters in our fold? Can they be brought into this discussion?

Although our discussion has centered on the taxonomy (the purpose of the discussion) I think there might be some functionality in del.icio.us that might require extension. Also, I have questions about the overall stability of the site, and its long term direction.

Is this a technology that will be shared (or is it available to share)? Do the site owners require additional resurces to make the site more robust?

There are certain functionalities I would like... importing and exporting bookmarks for instance. Also, group tagging (where you can tage several bookmarks at once with same tag). I am sure there are others.

I think nptech tag should be used for anything... websites, documents, products, people, etc. Anything with a URL.

This means that people will have to start coding their URLs with tags to document what type of resource it is.

My delicious entries code product/service, open source/commercial, etc, consistent with the coding in the advocacy technologies spreadsheet uploaded in the file system for this group.

If the nptech thing takes off, we can start talking about other codes like nptech_people nptech_product, etc. The actual entry in delicious would be "people" "product" etc, but there needs to be some concept of a taxonomy associated with the nptech space.

marnie webb said:

However, I'm ambivalent about the idea of only pointing to entire websites.

--

I'm not ambivalent, i think that nptech categorizations should apply to any url -- site home page or subpage or standalone doc or other should qualify provided it is relevant. Thanks for pointing this distinction out Marnie.

Michael Maranda said:

Regarding http://del.icio.us

Aside from some avid users of this bookmarking site/tool, do we have any of the site developers/proprieters in our fold? Can they be brought into this discussion?

Although our discussion has centered on the taxonomy (the purpose of the discussion) I think there might be some functionality in del.icio.us that might require extension. Also, I have questions about the overall stability of the site, and its long term direction.

Is this a technology that will be shared (or is it available to share)? Do the site owners require additional resurces to make the site more robust?

There are certain functionalities I would like... importing and exporting bookmarks for instance. Also, group tagging (where you can tage several bookmarks at once with same tag). I am sure there are others.

Michael,

You can find out more about del.icio.us on its about page. There is a pretty robust mailing list (it's linked from that page) which gives you access to people working on the tool. Because del.icio.us has an open api it's possible to get at the kind of information/functionality that you are talking about. I don't know of any del.icio.us add-ons that give you the group bookmarking functionality.

There isn't an explicit export function. I've bungled my way around this by periodically downloading the XML file associated with my del.icio.us account. I can open the resulting file in Excel (or the spreadsheet application of your choice) and can get at my data (links, tags, etc). I'm sure there is a more elegant way to do this but it seems to work just fine.

I think I get at the point of your questions: we are talking about using an application that we don't control (at all) for a project that could end up becoming kind of important to all of us and providing valuable information for seeding other relevant projects. Now, if I was consulting with a nonprofit organization I'd say don't do it. However, I think this tool gives us the power to jump start this effort before codifying it into some other application.

I guess that begs two questions: How are we going to get the word about this out to a broader audience and where will store the resulting data so that it can be used by anyone interested?

Hi I'm Lisa I'm a database person

I think the idea of a taxonomy is great as it can be used in drilling down in a search.

think

nptech

Software

OpenSource

Database

or

nptech

Software

Windows

Free

Database

This is important for database products where a "Google type search" isn't going to be fast enough or if the database system used doesn't have a text search built for it.

This also helps for classification purposes and cataloging.

Where some work has to be done, besides agreeing on the terms to use, is to decide if you allow the same word to appear at different levels of drill down or if you try to focus the words to specific levels of drilldown, meaning you have to prioritize the words ( Does OpenSource/Proprietary belong to a higher level than Operating System etc.)

The above would apply to an XML taxonomy as well since XML is a series of drilldowns in a tree structure (at least usually) for storing information

These is just some "thinking outloud" I'm not sure if it is useful, but it might be worth considering.

Lisa

marnie webb said:

The key, I think, will be a concentrated analysis of the collected data and then publication of the results.

Thinking out loud here. What do others think?

I'm not clear on what you meant by this statement.

And thank you for explicating what was behind my concerns about developing this project on top of delicious. If anyone is on the list with some of the developers, it might be nice to invite them here.

Is this the extent of the techfinder taxonomy, or is it deeper?

http://techsoup.org/techfinder/index.cfm?p=browse

If it is deeper, where can I find more detail?

That's it for Techfinder. It is purposely not deep, instead uses a faceted taxonomy to accomidate complexity... pick one or more from each category... a single item can therefore exist in many potential "heirachies"

actually, there are 6 top level facets in TechFinder: here's a view of this from a dev environment page.

http://dev.techsoup.org/techfinder_v1/unskin/index.cfm?p=browse_ng_2&tftaxid=68&tftaxid=8#b

David's right that it is a faceted approach, but TechFinder's taxonomy is designed to be deepened, for example you can see under Consulting in the above url, that 2 subelements, Organizational Development and Tech Planning are listed there. No deeper levels have been developed yet, and none have been suggested.

Another article that might of interest to folks here: Folksonomies - Cooperative Classification and Communication Through Shared Metadata. I haven't actually read it yet...

update: okay, now I've read it. Really, it's an explanation on why the kind of thing we are talking about here may be a good idea. Interesting.

There is some interesting discussion of topic maps:

http://www.omidyar.net/group/topicmaps/news/1/

I think their might be some fruit in linking between these dialogues.

So, do we think that we should push this out to a broader group? Maybe the folks on the Information Systems listserv as a start?

marnie webb said:

So, do we think that we should push this out to a broader group? Maybe the folks on the Information Systems listserv as a start?

--- I think yes. I was thinking we could wait til we have a time set aside to build up the list first, but to invite people now with the understanding this is an initial step seems fine. I'm out and tied up most of Jan, so I won't be chiming in much. Marnie-- if you can be sure to follow up on any questions that folks would have on the IS listserv, then going ahead might be a good way to test the waters.

Phil, I can keep an eye on it an fend any replies.

What do other folks think?

Sounds great. Is sub. info for that list available in one of our resources folders?

Here it is:

http://groups.yahoo.com/group/Information_Systems_Forum/

I just started using del.icio.us and I am interested in getting involved with this.

I am not sure I understand what this is for. How is this different then techfinder?

I really like the idea of including articles and pictures. I think this would be a great way for nonprofits to start writing blogs and using rss feeds.

How many people know about del.icio.us? I know I found about that site through these discussions. I am not sure how many people you want contributing to the tag list, but an idea might be to write something about del.icio.us on Techsoup and/or Information Systems group.

Also, the subscription feature does not work right now on there.

Emily,

Thanks for your questions.

Let's see. You asked:

How is this different then techfinder?

Well, I think the prime difference is that this can capture a greater amount of information that TechFinder does currently. TechFinder focuses on people; this can capture anything that the participants believe to have to do with nonprofits.

The basic idea is that this will:

1. Give us an idea of the kinds of information people need. This is based on the assumption that participants are using del.icio.us (at least at some level) for their own self-interest. That is, they are bookmarking things that they need and would like to be able to find again.
2. Give us a natural language for describing the items captured.

I don't know that many people know about del.icio.us at this point. What I'm planning on doing is writing about this experiment and posting it to the Information Systems mail list and perhaps putting up a brief message board post on TechSoup.

In addition, participants, if it's valuable to them, can promote the effort to their own constituencies through weblogs or whatever forums they have available to them.

You're right; the inbox is not functional right now. However, using an RSS/Atom web feed aggregator you can subscribe to:

http://del.icio.us/rss/tag/nptech

This points to the RSS feed for the tag "nptech."

I think another key next step is to write a brief del.icio.us tutorial aimed at nonprofit users. It would be great to get some other people -- people who are not necessarily alpha geeks but are tech savvy involved in this.

I'm thinking of reaching out the the members of the Alliance for Nonprofit Management via their listserv for some additional participation.

I think I understand the difference now. I am going to start looking for links to add. I spend alot of time searching the Internet and I know I will enjoy finding more resources to add to the nptech list. Since I am a recent college graduate, my bookmarks with the nptech tag will probably not be something that I would need. I would be adding bookmarks that I think would interest nonprofits.

When you write about this, you might want to mention about subscribing to the feed for the nptech tag. I just added it to my feed list and the tag is listed in my bookmarks on del.icio.us.

Even though I am still learning del.icio.us, I would be interested in helping with a tutorial. How do you want tech savvy people involved in this?

I think the best way that tech savvy folks can be involved in this is, quite simply, to highlight interesting projects. Of course, if people wish to use the extended field in del.icio.us to explain why the bookmarked URL is of interest -- mores the better.

Thanks, alot, Emily, for participating. Points to ya.

Well, I've started some outreach about this. You can see it here: Time to get serious about the Nonprofit Technology Taxonomy experiment.

Okay, so there are quite a few bookmarks and participants joining in on the nptech tagging project. As pointed out by a few commenters, it will be easy enough to extend this to flickr and technorati tags.

I've started thinking about how to describe this concept and push it out further than the relatively geeky audience that's participating at this point. I'd love some help in think about the potential benefits. What can this be used for?

I was just starting to come up with ideas for the benefits of this.

I was thinking this could be a start for a non-profit technology virtual library.

I have been trying to tag things that are blogs, blog entries, news articles. I started thinking how great it would be to start a collection of these.

There are also many companies and organizations that are listed with the nptech tag. Those can be organized by their locations. I am not really sure how this part would be different then techfinder, but I am sure we will think of something.

Thanks, Emily. Those are good uses for the bookmarks -- and they can inform existing efforts like TechFinder.

I'm also wondering what the taxonomy itself can be used for?

marnie webb said:

I'm also wondering what the taxonomy itself can be used for?

Are you referring to the nptech tag page on del.icio.us or the tags for the webpages?

I was thinking the collection would be in a form of a web directory and/or database with the tags.

I'm referring to the actual tag words.

So here's my thoughts on the potential outputs/outcomes for this project:

• A collection of bookmarks. As Emily's noted, this project has resulted in 400+ bookmarks. This can provide a basis for people delving into the topic of nonprofit technology.
• A collection of interested users. From my standpoint, this is also interesting. Via the users, I've found weblogs and projects that I didn't know existed. It feels like, at a base level, the start and/or extension of a community.
• Weigthed importance of links. This is very rough but the number of users who link to any given URL provide a "weight" for this URL. I'm reluctant to translate this directly to "importance" but it's a judge of reach, need, and interest.
• A collection of words. This one requires more of a leap and I'm not sure if I'm making the right leap or explaining the jump very well. So, if we pull together all the words that different users used to tag a single link, we'll get an interesting collection of descriptive terms. Some of those words will have been used more than once. Those words become highlighted -- these are the words users (even if they are relatively geeky users) are using to describe a parrticular link. Begin to link those together and this can provide the hierarchy for the directory that Emily talked about. It can also begin to provide the basis of a nonprofit taxonomy.

The real question for me is about that last one. How realistic is that? Can a folksonomy be mined to begin a taxonomy? And what, exactly, are the uses of this taxonomy?

Getting at the data in the nptech tag isn't as easy I thought it would be.

After poking around and convincing myself I couldn't find the answer on the web, I posted a request for tag analysis help to the del.icio.us discussion list. I wanted the helped I'd asked for here.

I found that I couldn't get what I wanted. At least not in some highly elegant organized way. I can port the html of the nptech pages into CSV and go from there with some initial analysis.

But I got a lot more.

Through the discussion (I have 30 emails in the thread though some of them were off-list), I learned a lot about a big flow in the experiment and ways to think about getting past to that flaw.*

The flaw, at least as I read it, is the relative obscurity and arbitrary nature of the "nptech" tag excludes any users who don't find out about the tag directly (via listserv or blog posts, for example) or who don't tag surf to it and then begin to use it.

So the initial project goals:

1. collect bookmarks having to do with nonprofit technology;
2. find unknown users;
3. provide a weighting, represented by the number of users bookmarking an item, for bookmarks; and,
4. collect descriptive words to inform a more formal taxonomy project.

are undermined by the exclusive nature of the proposed unifying tag.

This leads me to the conclusion that, once some tag analysis has been done, we should do some analysis on the most bookmarked URLs to stretch beyond the users who use "nptech." That will certainly inform the first three goals. But it will really inform that last. We will be able to collect additional words and word groups for the taxonomy process.

Today was, in my book, an absolutely fabulous learning day.

*It's my policy not to quote anything that I haven't been given express permission to quote and/or can't be linked to on the web.

I was actually trying to find some information on it too. I was thinking there would be a way to take the html and convert it to xml, and then work with the data from there. I know the basics of this so I don't know if that would work.

I read the responses to your request. I started looking at other users lists if they had a link that was on nptech. I found some more links to add from that. It did not occur to me to add + to get more to add. I knew about using + but I did not know what exactly it did. I learned that today from this experiment.

Now if we can figure out a way to do the following:

1. tell others about nptech
2. teach others how to use the site so they can add links
3. find out who exactly the users are (not just by their username, but a way to e-mail them)
4. tell users who are participating to bookmark links that they like that are already on nptech - even if they don't use the nptech tag to bookmark it (I know I was confused about this one)

I think we still have more bookmarks to add and find others who are intrested in this.

I think we'll try mining the keywords used in searches of techfinder to find candidates for taxonomy entries; looking for the most commonly repeated searches. I bet there's also good stuff to mine in the keyword search values entered in the techsoup search.

Phil Klein said:

I think we'll try mining the keywords used in searches of techfinder to find candidates for taxonomy entries; looking for the most commonly repeated searches. I bet there's also good stuff to mine in the keyword search values entered in the techsoup search.

You superstar you. ;)

This is a fantastic idea. Who has the capability to start storing these in a data base? I volunteer to do the analysis if someone sends me a csv.

How will you get the information for the most commonly repeated searches? I was looking at some of the messages at the techsoup forums to add as bookmarks.

I noticed the keywords/tags I was using alot for my bookmarks were (besides nptech, nonprofit, and technology) email, fundraising, advocacy, online, webdesign, webhost, marketing, communication. I am sure there were some others but I started to add those terms to a type of nonprofit, like environment to see if I find anything from that.

Would it be possible to add a message or something on techsoup about the nptech tag?

Emily W. said:

Would it be possible to add a message or something on techsoup about the nptech tag?

I have a bit of an in there. I suspect I could get something put up about it ;) In the meantime, I can certainly post something on the community boards about the nptech experiment and then see if that can bumped onto the front page.

I don't know, though, how many of the TechSoup engaged user base are also del.icio.us users. I'll also post some stuff about del.icio.us in general.

David,

I got a couple of offers of help on getting the data out of the del.icio.us pages from folks on the del.icio.us list. As soon as I get the data, I'll up load it as a CSV and attach it as a file here in Omidyar.

I'll also see if we can get some search data out of the TechSoup search logs to add to the data pile.

I am trying to figure out what I can do from here. I am still adding bookmarks to nptech, but is there something else I can be doing?

I can try to help with the data or just spread the word about the tagging.

Emily,

Keep adding and spreading the good word. The folks that offered assistance are dumping the data on my so I'll be making those files available as I get them.

Once that happens, I think there's some room to talk about how to revamp the project, etc.

:mw

Moving this into a new thread: NPTech Experiment.