kentbye's picture

Fractals & Folksonomies: A New Map for Participatory Journalism

| | | |

The press is supposed to provide us maps that help us understand our complex and chaotic world, but the media's maps have failed to keep pace with the exponential growth of technological innovation and changes in our society.

Below I propose that folksonomies and playlisted sound bites could be used to map out the links and associations between the Long Tail of factual nuggets that are usually lost on the cutting room floors of news stories or documentaries -- as well as how "fractal geometry" can provide a powerful metaphor for helping visualize and comprehend this complexity. This type of approch could be extended to journalism and other knowledge management contexts.

Participatory journalism is the key to mapping out this fractal-like map of interconnections, and this post describes what "news as a conversation" might look when it is scaled up beyond the linear limitations of blog dialogues.

Continue Reading...

kentbye's picture

My tagroll


I've been using for a while to track the websites that cross the threshold of my attention stream -- a lot of these sites have to do with researching technologial issues and political issues surrounding the Echo Chamber Project.

I thought that this latest tagroll feature would be a good time to show a snapshot of my brain.

I'm also phasing over to feedburner feeds where it sends my latest tagged stories out through my RSS Feed -- which is the feed that will also contain my latest blog posts, video blogs, and all podcasts (both interview audio and community/technology audio).

kentbye's picture

Playlists are to Music as Edit Decision Lists are to Film

| | | | | | |

When the timelines of edited film sequences are exported, then they are flattened into an "Edit Decision List" that is somewhat analogous to a musical playlist and an academic syllabus or H20 playlist.

UPDATE: I explore the how this playlist concept can be applied to filmmaking in conversations with Lucas Gonze, Colin Brumelle and Farsheed

JD Lasica just posted a video interview with Molly Krause of Harvard's H20 playlist project.

You can think of H20 as a way to share a college class syllabus. It's an ordered reading list that can be used to aggregate knowledge from experts. They describe it as an "open source, educational platform that explores powerful ways to connect professors, students, and researchers online."

Here's an example of a H20 reading list that should give you an introduction to "Social Bookmarking with" written by Brian Del Vecchio.

H20 tracks derivatives made from playlists as a way to track the relative authority, expertise and reputation of a given author -- much in the same way that academic citations in peer review journals are a way to measure these same metrics. But the H20 playlist format decentralizes this process from the normal gatekeepers and allows for a much more grassroots and bottom-up approach to this concept.

So as Krause says in the interview, you can think of these playlists as a way to provide guided maps to particular fields of study.

My understanding is that playlists have gained a lot of popularity because it is a way for people to create sequences of songs to play on their computer or mobile devices. Because more and more individual songs are being digitally distributed and separated by the order in which they usually play on an entire music album, then playlists have been able to recreate these musical experiences much in the same way that DJs have done.

So Harvard has expanded this playlist concept from music to academic information, and I would like to expand it even further to a journalistic and filmmaking context.

Netflix is already using the playlist concept for distribution of DVDs with their "Netflix Queue." You select videos that you want to see, and then you determine the order in which you receive them.

This can be extended to the actual generation of films because filmmakers are essentially doing the same thing except with multiple video and audio dimensions synchronized by timelines and smaller nuggets of information (i.e. a sound bite vs. an entire DVD).

When the timelines of edited film sequences are exported, then they are flattened into an "Edit Decision List" that is analogous to a musical playlist and an academic syllabus or H20 playlist.

Edit Decision Lists can be generated with a web browser interface, and then dynamically translated into online edits by using the SMIL open standards -- or into offline edits by using Final Cut Pro XML interface that I've described before. I've been able to successfully accopmlish both of these in the tests that I've done.

Most people get completely lost by this point, but I'm basically exploring the idea of using playlists for the collaborative generation of media much in the same way that Harvard is exploring playlists for the collaborative distribution of knowledge.

I was very happy to discover that H20 backend has been open sourced, however the code was a bit too complex for me to parse.

But I'd love to catalyze an effort to port some of these concepts from H20, and into Drupal.

I've been in contact with the two Drupal developers of the playlist module, and I hope to talk to them more about it soon.

I also happened to meet "playlist maven" Lucas Gonze of at the Open Media Developers Summit, and may pick his brain about the function and culture around playlists -- as well as best practices for tracking related and derivative playlists.

So with that, I'll share the e-mail and comment below that I just sent off to's JD Lasica (whom I also had the chance to meet at the summit)...

kentbye's picture

Screenshots of User Interface for Distributed Editing

| | | | |

I have some preliminary screenshots for what the volunteers will see when they help order sound bites into sequences.

This has been some of my first Drupal development, and I'm sure that this interface will continue to evolve -- but I just want to show what I have so far.

More below...

kentbye's picture

Progress on Collaborative Filmmaking Infrastructure

| | | | | |

In order to have more volunteers get involved with this project, then I need to build the infrastructure in order to put them to work in helping edit the film.

Lately, I've been able to make some promising advances and gain some new insights on my original plan for Collaborative Filmmaking. I completed some milestones that have allowed me to prototype the workflow and make alterations to the original flowchart.

Once I get these tools into place, then a lot of my writing will become less technical and geared more towards creating videos that will instruct others how to participate.

This is what my plan for collaborative media currently looks like:

Collaborative Filmmaking Flowchart Version 2.0

There are more details below...

kentbye's picture

Database Questions for Drupal Tagging by Other Users

| |

I just e-mailed the following questions to the TagDB Listserve

I'm hoping to get some insights for what it'd take to add additional capability to Drupal where users could add folksonomy tags to other nodes that the did not create.

Hey everyone.

Currently Drupal can add folksonomy tags to the node that a single author creates, but it's not yet possible to have other users internally add tags to other nodes created by other users.

I have a database question for specifying a special node-type in Drupal for tagging folksonomy data by other users.

I think that it'd probably be possible to do this using the "flexinode module," but I'm not sure if it would provide scalability problems since it spaces out the metadata across so many rows.

So I thought I'd give an example of what I see the current implementation would look like with flexinode, and ask the following questions:

Would the following setup work?
What would a better table setup be?
Is the way that Drupal tracks tags the most optimized and scalable?
What are other tables for indexing would be useful?

Here's a summary of third-party Drupal tagging, with a more detailed example below:
NODE: nid / type / title / uid -- [The Tagged Object, blog/page/flexinode/etc. type, e.g. nid = 624]

NODE: nid / type / title / uid -- [The Tag Node, "Tag Another Node" flexinode type, e.g. nid = 599]
TERM_NODE: nid / tid -- [1st folksonomy tag]
TERM_NODE: nid / tid -- [2nd folksonomy tag]
TERM_NODE: nid / tid -- [3rd folksonomy tag]
TERM_DATA: tid / vid / name -- [1st folksonomy tag]
TERM_DATA: tid / vid / name -- [2nd folksonomy tag]
TERM_DATA: tid / vid / name -- [3rd folksonomy tag]
FLEXINODE_DATA: nid / field_id / textual_data / numeric_data / serialized_data -- [URL of object being tagged]
FLEXINODE_DATA: nid / field_id / textual_data / numeric_data / serialized_data -- [Additional Comments]

kentbye's picture

Progress on Phase 01

| | |

Part of my Phase 01 vision for this project was to have a way to collect feedback from volunteers on the different interview sound bites.

I've been able to successfully upload all of the sound bites from the Bill Plante interview into Drupal.
You can check it out here:

At the moment, only I'm able to add folksonomy tags to these sound bites, but I'm looking at ways for volunteers to tag these sound bites as well. Currently, there isn't an existing solution, and I'm asking around it anyone has some good ideas.

kentbye's picture

Echo Chamber Project Vlog Episode 1: Introduction

| | | | | | |

Introducing the first Echo Chamber Project video blog entry & vlog!

Description: First vlog episode about an open source, investigative documentary on how the television news became an uncritical echo chamber to the countdown towards war in Iraq -- and proposed tools for collaborative journalism that can provide some solutions.

Featuring: Jay Rosen, Dan Gillmor, Doc Searls, Jonathan Landay, Pamela Hess, Bill Plante, Halley Suitt, Marilyn Schlitz, Kent Bye and 60 others.

To Watch the Video click here -- or on the picture below -- or try here if that link doesn't work. Check back in 10-15 minutes if neither work, the Internet Archive has been a bit spotty.

Sit back, relax and enjoy the show!


(6:15 minutes / 15 MB)
Download QuickTime

Listed below is

* A full transcript of this video with additional links
* How to keep informed with the project (Vlog RSS / Blog RSS)
* How to get more involved
* Click here & scroll to the bottom to leave feedback or other comments.

kentbye's picture

Collaborative Filmmaking Flowchart

| | | |

I've completed a first draft of my first video blogging entry, and I realized that I needed a graphic to talk through just exactly how I plan on doing Collaborative Filmmaking by using Drupal and Final Cut Pro XML.

Here is the graphic that describes my vision of the Phase 01 infrastructure needed to collect folksonomy and quantitative metadata from volunteers on the soundbites:

Collaborative Filmmaking Flowchart

It's a bit difficult to convey everything in a soundbite or two, and so my second videoblogging entry may go into more details about it.

But for now, I'll just post this graphic online in a few places, and provide a brief overview and hook in my first vlog entry.

kentbye's picture

Joining a Folksonomy Community of Practice

| | | | | |

After a warm-up specification of Drupal functionality, I've started to map out the user interface that I need in order to productively collect collaborative input from volunteers, and then feed this gathered intelligence into the editing process.

I've discussed the mechanisms for this Phase 01 of development as being folksonomies, tag clouds, Drupal & Final Cut Pro XML.

Drupal already has some folksonomy capabilities, and the open source nature of the software allows it to be altered by computer programmers familiar with the open source scripting language of PHP, the open source database of MySQL or the inner workings of Drupal.

I've been interacting with a number of developers over the last couple of days, and I've found some potential leads for some help.

I also revisited Phillip Keller's post about the technical database details for the database schemas that are fueling social bookmarking tagging sites such as or flickr. Keller provided a lot of interesting insights for how to make Drupal's folksonomy implementation more scalable for the types of features that I wanted to include.

Keller's post led me to the mailing list, as well as to Nitin Borwanker's interesting tagschema blog that contains philosophical musings about the systems level database design issues for the flood of data technologists need to help manage in the 21st Century. Borwanker describes it with the metaphor of the 1 Billion Row Problem -- referring to the daunting task for managing a database that has 1+ billion rows.

On Tuesday, Borwanker announced that he's starting a special spin-off listserve from the list in order to broaden the conversation to other developing tag projects, and to brainstorm the best practices for designing the backend databases for large-scale tagging websites.

I joined this Folksonomy Community of Practice with the intent of finding some help in getting some feedback on some of my a href="">Phase 01 ideas and to get some help in getting them rolling.

Below is the introductory e-mail that I sent out the listserve this morning. You can monitor the responses from a distance -- or sign up to the folksonomy discussion list yourself.

Syndicate content