State Of The Art/Modelling knowledge artifacts

From LiquidPubWiki

Jump to: navigation, search

Contents

Art, Poetry and other pre-information age artifacts

Scientific community and publication sites

Early attempts

  • ETAI - http://www.ida.liu.se/ext/etai/
    Electronic transactions on Artificial Intelligence: this is an Electronic Journal that tried to operate like a normal Journal (with papers submitted and approved/rejected by a board) but that ALSO included some of the funcionalities of Internet (discussion, etc).
    Ultimately ETAI did not succeed in drawing the people in by presenting “significant advantages” over the normal journals. The way the interaction is presented is too rigid for the often chaotic creation and discussion process.

Open Access Initiative

The Open Access Initiative wants to give full-text online access for any user to research articles published in peer-reviewed journal There are two ways that an article can now be published under Open Access, the first is self archiving and the second is submiting it to an Open Access journal.

“Scholars are paid by research funders and/or their universities to do research; the published article is the report of the work they have done, rather than an item for commercial gain. The more the article is used, cited, applied and built upon, the better for research as well as for the researcher's career”

www.unesco.de/openaccess-en.html

Current Publication Sites

Web 2.0

Collaborative authorship and Social Web

  • Wikipedia - www.wikipedia.com
  • Wikis are generally designed with the philosophy of making it easy to correct mistakes, rather than making it difficult to make them. Instead of trying to prevent malicious modifications it makes easy to undo them. This openness has greatly helped propel its growth.
  • Wikipedia does not take credentials into consideration and for that reason it heavily favors consensus and popular information.
  • Much of the coordination of the editing of Wikipedia takes place on the "Talk" pages associated with each individual article. Assessment of the quality of articles based on reviews from the administrators
  • However Wikipedia requirements (well-know subjects, no new research, neutral point of view) leaves the door open for our research, which will mostly deal with exactly the opposite (new material, heavy on opinion)
  • Another limitation of wikis, in general, is their linear history, also big changes are much more likely to be resisted by other users than small progressive ones.
  • Ylvi - A Semantic Multimedia Wiki. Ylvi has a focus on the administration of multimedia data and supports fine grained access control (RBAC) of its resources. It is based on the multimedia framework METIS and makes use of its generic data model and its sophisticated plug-in infrastructure. See the survey paper at: [1]

Basically a wiki but with locking mechanisms. Meaning once one creates a page can limit its edition from others. Swiki is a good example of a platform for collaborative efforts(software development) and applied manuals.

  • Ning – www.ning.com
  • Knol – knol.google.com

Semantics based on community input

  • Flickr - http://www.flickr.com/
    Photo upload site/community. Flicker uses folksonomy to tag images. The system itself seems to be a simple tagging without weight, original user can set the tags and the user can suggest new ones. Uses a flagging system to quickly take out of circulation inappropriate material.
    Tag disambiguation is done by associating tags into Tag Clusters. Normally just by seeing the other tags in the same cluster, you will be able to tell exactly which meaning that the tag is referring to.
  • Del.icio.us - http://del.icio.us/
    Bookmark sharing page. Meaning for tags can be filled by users when missing. non-hierarchical keyword categorization. No disambiguation is done, it is expected that users use multiple tags at the same time to disambiguate themselves.
  • Digg - http://digg.com/
    Similar to del.icio.us but regarding to news. Users can digg news and the ones with more diggs get advertised on the front page. There are sub clasification options for the types of news and Friendlist features.

Software Versioning and Development

Version control systems

Distributed Revision Control Systems

Online Software Development Services

  • Sourceforge - http://sourceforge.net/
    One of the most popular open software collaborative development sites. It offers public repositories along with a series of extras like automatic project domain (http://project-name.sf.net.), bug trackers, request lists, forums, mailing-lists, infrastructure to receive donations, blog, calendars, documentation, WIKI, etc
    For creation of new projects, Sourceforge needs the intervention of a 'review staff'. Along with the basic info for your project, you are demanded to enter a: “a detailed, accurate, English-language description of your project, including technical aspects which cause it to differ from similar, existing solutions”, this will be used to avail if the project indeed is worthy of being hosted in the site. After submitting, further input from you may be needed to justify the presence of your project.
    The next step is giving the project its tags or its position in the hierarchy of categories. Five complete tree entries are mandatory to continue
  • Ohloh - http://www.ohloh.net/\
    Sourceforge-like development site, but it also crawls the projects' repositories to find out the level of activity of the project and its developers along with other misc info (licence conflicts). It counts commits and 'kudos' given by other users to rate authors.
  • Darcs - http://darcs.net/
  • Launchpad - http://launchpad.net/
    Used by the Ubuntu Linux project and others. Reportedly to be released as free/open source software later this year.

Document management and Versioning

Document management

Project/Conference management sites

Research repositories

Data and Knowledge management

Markup and Metadata

Knowledge discovery and sharing

Web 3.0 or Semantic Web

An example of this is the Fluid Web Approach where content has a great deal of metadata describing it and relating it with other content. Furthermore, each unit in this approach has not only information on the nature of the data it contains, but also on the way to execute it by association with other units. This enables to the sharing and versioning of content among several different applications.

Search and Navigation

Web search and desktop search

  • Citeseer - http://citeseerx.ist.psu.edu/
    Website that keeps citations of publicly available papers along with their abstract and full text. It presents citation rank and graphs.
  • DataBase systems and Logic Programming - http://dblp.uni-trier.de/
    Similar to Citeseer but this includes only abstract and citation from the papers. This allows it to include non-free papers. Its data was loaded by hand.
  • Google Scholar – http://scholar.google.com
    It has citation indexes and present related works to a paper (based on how similar are the papers compared and taking into account the relevance of each paper). Google Scholars keeps closed the number of the journals they include and the update rythym but its performance is normally compared to that of subscription similar services.

Semantic Search

Navigation of related elements based on Metadata

Retrieved from "http://wiki.liquidpub.org/mediawiki/index.php/State_Of_The_Art/Modelling_knowledge_artifacts"
Personal tools