Navigation:
Documentation
Archive



Page Tree:

Child pages
  • Fleshing out ideas for Demonstrators

This wiki space contains archival documentation of Project Bamboo, April 2008 - March 2013.

Skip to end of metadata
Go to start of metadata

On this page, please add your ideas for demonstrators and your comments and suggested elaborations about potential demonstrators listed. Right now these are just ideas for Demonstrators -- we need to recommend which ones should be built.Kaylea Champion is staff liaison for this working group and also  the PM for Demonstrators for the Project. Feel free to get in touch directly with Kaylea if you have questions, comments, etc. for the bamboo program team regarding Demonstrators.

Ideas for demonstrators from Workshop #2 brainstorming

Demonstrators display a concept, serve as a reference model, and reflect and test the results of discussion and analysis. For more, see About Demonstrators.

In thinking about creating Demonstrators for Workshop #3, we want to identify projects that fit the size rule of 3 -- i.e.,  demonstrations that can be developed into presentable form in 3 hours, 3 days, or 3 weeks. Demonstrators (typically) will not be fully functional; they may in fact just be mock ups or streaming media designed to illustrate a model or approach useful in discussions about Project Bamboo. We have access to some Project staff to support construction of Demonstrators, including systems, programming, and visualization support, but in many cases the best people to work on the demonstrator are the people who already know the tools, disciplines, etc., so if you have something to contribute, please say so. See also information on the Join Us in Demonstrator Work page.

The following is a list of Demonstrator ideas surfaced at Workshop #2. Please add new ideas and your comments on the ideas listed here. Add directly to the list and/or by creating comment threads at the bottom of the page. For more about using this wiki see 1, 2, and 3.

  1. A demonstrator that illustrates the idea of turning user behaviors into criteria of relevance.
    1. It would combines social networking with social discovery and coediting, across disciplinary and knowledge domains. It would also allows for student-teacher-researcher interaction.
    2. Purdue University is working on one and its participating team would be glad to give a presentation/demo/workshop.
  2. An rss feed ecosystem which would illustrate the idea of collaborative collation and distribution of information (a la Wordpress... We should not ignore the obvious).
    1. Sorin Matei at Purdue has been experimenting with such applications for some time and would be happy to give a demonstration.
  3. A demonstrator of an integrated, collaborative learning and discovery system that funnels information tagged on the basis of time and space criteria to the user wherever the user might be (real or virtual representation of geographic space). The platform is fully functional, including location aware capabilities (information is delivered on the fly when the user is at a specific location). In addition, the platform can be used for historical reenactments and for connecting traditional (analog) digital realms of information. see http://visiblepast.net
  4. A demonstrator wiki that would allow us to collect the items for a tools registry (entries could be tagged by the domain/activity/themes/degree of interoperability).
  5. A demonstrator that would create a template for use cases and case studies template that will allow us to define the typical problem - not just a list of tools, but what are they trying to solve? Can they be reused, reconfigured in a specific scholarly process?
  6. A generic workflow representation for scholarly activity which would be customizable by domain
  7. A demonstrator service to get an image that facilitates sharing of content across multiple formats
  8. A demonstrator of a Template for defining problem sets
  9. A demonstrator to facilitate sharing text or interoperating across shared texts; e.g., sharing multiple texts from multiple archives and asking an analytical question across all of them
  10. A demonstrator to represent discovery of content as well as tool discovery registry.
  11. A demonstrator service to extract entities: extracting people, locations, other things for which people can specify ontologies
  12. A demonstrator service that maps extracted date entities to MIT timeline (so as to exploit SIMILE), and extracted location entities to Google Maps
  13. A demonstrator to pull annotations across repositories
  14. Scholarly mashup environment - stitch together multiple tasks

8 Comments

  1. Unknown User (t-cole3@uiuc.edu)

    We might be able to implement a variant of idea #4 across some real image resources by exploiting Djatoka, an image processing service created by Los Alamos National Labs Library (Herbert Van de Sompel and Ryan Chute). This service can ingest images in a range of standard formats as long as they're accessible by URL and allow users (or other applications) to maniputlate them -- selecting sub-regions of the image to view, zooming out to different resolutions, rotating the image, etc. Zoomed, rotated, cropped images are cached and can be made available for later re-use at a persistent URL (using a scheme based on OpenURL). For more see recent D-Lib Magazine article (http://www.dlib.org/dlib/september08/chute/09chute.html). The idea would be to show potential to implement standards-based image manipulation tools (Djatoka is based on JPEG2000 standard and a number of standard image manipulation libraries).

    UIUC would be willing to help implement this a demonstrator over content from the Digital Library Federation's Aquifer-American Social History Online portal. We've already had conversations with Djatoka's developers who have expressed a willingness to help create a demonstrator along these lines. Could probably implement over something like Flickr Commons as well.

    Other suggestions for target content or elaborations on this idea for a demonstrator?

  2. Unknown User (smatei)

    I am not clear if this is a request for demonstrators or a list of possible demonstrations that can be used in the Seminar Three. Can you please add relevant language at the top of the page to explain this. Also, throughout this site, we need to add information about the relevance, utility and editorial process (what to do next) for each page.

  3. Unknown User (martinmueller@northwestern.edu)

    Here is a 'demonstrator' that may be relevant to Bamboo but doesn't depend on it and can't really make use of the offered help. Katrina Fenlon, a graduate student at UIUC's Information and Library Science school, will do a practicum with Tim Cole, Alan Renear, and me in early 2009. The Rare Book Library at UIUC is thinking of digitizing its valuable collection of 19th century British novels. Can the resultant OCR texts be transformed with an acceptable level of human labor into TEI texts that will play well with existing digitized collections, such as Early American Fiction or the TCP EEBO, ECCO, and EVANS projects? If the answer to that question is positive, one can see a five- to seven-year project in which high-value texts in the public domain become part of a very large and fully interoperable collection of texts that will form a single document space for many inquiries. The approximately 50,000 texts from the TCP collections will pass into the public domain after the middle of the next decade. Add a large collection of similarly encoded 19th century texts and you have what you can call a Book of English or cultural genome of written English from 1475, the date of the earliest printed book in English, and 1923, the current copyright cutoff.

    Katrina will work with a handful of 19th century novels. She will look at an XML output in which the very mechanical XML generated by the OCR process has been transformed into a TEI format, which will require various forms of adjustment. She will look for the necessary adjustments, keep a careful log of what needs doing, and think about repetitive steps that may lend themselves to algorithmic treatment. On the basis of her experience we will try to figure out realistic estimates for human editorial labor and think about a model of distributed editing that would allow for a user-driven process of adding to the Book of English. There are terrific opportunities for volunteer labor in this field, provided one can construct a sufficiently user-friendly and network-based editing environment.

    I should add that in the Monk project we have solved the problem of transforming TCP texts (and similar texts) into a linguistic corpus where a single morphosyntactic tagging and lemmatization scheme has been applied to texts from the late 1400's to the early 1900's, thus making those texts fully comparable at the level of their metadata. If Katrina's experiment gives us hope that OCR texts can with acceptable human intervention be turned into good enough TEI editions, these editions can without trouble be morphosyntactically tagged and become part of a Book of English.

    That is a lot of interoperable content.

  4. Unknown User (patrick_yott@brown.edu)

    I think that in able to do something meaningful in the limited time available, we should focus on Tim's suggestion of using the Djakota tool to build an image service provider.  It could be extended, without much trouble (I think) into an image annotation service as well, given the ability to reference coordinates in a source image...

    p

  5. Unknown User (ray@ischool.berkeley.edu)

    I have been meaning to let people know about some already existing "demonstrators" -- prototypes of systems

    that my research group (and former students) have been working on in this area. The

    first set is for a pair of projects working on support for scholars for biographical information and for Irish Studies.

    the demos page includes links to the demos themselves and to a couple of videos on what they do.

    http://metadata.berkeley.edu/demos/ 

    The second is using Djatoka and providing images of a medieval french manuscript with transcription overlays

    as you move the mouse over lines (and variations). I have found that there are sometimes layout problems

    if you aren't using Firefox with this demo. (This one is the work of Rob Sanderson at the University of

    Liverpool, using his transcription of Froissart's Chronicle that he did for his PhD thesis along with 

    designing the markup and search capabilities for the transcription)

     http://www.openannotation.org/adore-djatoka/

  6. Unknown User (smatei)

    Hi, I would like to reiterate the idea of considering the Thought Ark platform (http://thoughtark.com) as a possible demonstrator. Thought Ark is a complete resource selection/management/publication system which recycles user behaviors automatically and within the normal flow of research activities for discerning between resources (mainly texts) that are more or less valuable. The system is open source and interoperable, meaning that it can be installed by any institution and the resources can be made available across institutions. In addition, the value criteria can be transferred across knowledge domains.

    1. Unknown User (khc@uchicago.edu)

      The scholarly networking group is looking at Thought Ark as well.

  7. Unknown User (markj)

    Thanks to Tim for inviting me to introduce another possible demonstrator.

    I gave a presentation about this at the Chicago meeting of Bamboo #1, and have discussed it in the Education working group. Kaylea has been in touch about developing it further, and Tim asked me to mention it here. We can discuss it on Friday, if there is time.

    The general idea is to develop a (multi)disciplinary applied wiki derived from Media Studies, nominally called The Media Ecology wiki. (This presumes that wikis are the best format to consider...)

    The sample focus project would be to seek out (and advocate for) streaming historical TV newscasts as objects of study. Two significant goals would be 1) to underscore the significance of television news to the goal of an informed citizenry, and 2) to better understand the role of television news regarding popular memory.

    The model we have discussed is a web-portal integrating web archives and analytical tools and e-publishing, in consultation with applied multi-disciplinary organization(s) such as the Association of Moving Image Archivists. I attended the recent AMIA convention, and this idea was met with enthusiasm.

    One take on the process to follow would be to locate historical newscasts already online; advocate for additional newscasts to be placed online; cultivate tool sets for tagging and metadata; investigate possible tools for precise citation and other scholarly notation; deliberate about possible licenses for mash-up capabilities, etc.

    This project would require skills across the full complement of Bamboo's constituency: Humanities, IT, and library personnel. Because of recent amendments to Section 108(f)(3) of the copyright law, there may be a key role for libraries as points of distribution.

    It is assumed that we might start small, working with local and regional collections, and aspire to work with major network news collections. (Could we eventually posit, for example, an Academic Hulu that provides fair use scholarly access to news libraries, etc?)

    One relevant IT process that was mentioned as a possible model for the more broadly conceived Media Ecology site is nanoHUB (https://www.nanohub.org/home).

    I look forward to your ideas and comments.