At CNI:

Social Bookmarking in an Academic Environment

Michael Winkler (Director, Information Technologies & Digital Development @ University of Pennsylvania) talked about Social Bookmarking in an Academic Environment. He introduced penTags, an application they had created to support students' development of annotated bibliographies in an cinema studies course, that's since been used to create research guides for the library and image portfolios and to support other personal uses of tagging.

pennTags is a relatively simple [3 table] application that integrates identity management (though the universities identity servers), tags (and annotations) and web resources. While much of the application looks a lot like del.icio.us or dogear, a couple of  things stood out to me.

- Genres can be assigned to posts: a review  could  be distinguished from a bibliography entry, from something else. This acknowledges that our purposes and motiviations in tagging aren't always the same.

- Types of annotations were distinguished:  two separate kinds of note fields were provided, so that 'formal' annotations (Descriptions) for a bibliography could be distinguished from comments (Notes) about the site. This acknowe

- Posts could be linked to Projects: the relevant parts of particular entries created for a particular assignment could be linked to a larger whole (through a "part of a bibliography" element) and output in a form that 'looked like' an annotated bibliography

pennTags seemed to push the boundaries of tagging for me; it did more than simple recording of words to aid discover (or re-discovery) -- it provided a way to annotate, index and cluster web resources, to retrospectively apply personally meaningful constructs.

/jt



My notes from the session:



Why?

- people make lists

-grocerylists.org

-itunes xhtml playlists

-reputation

-lists as way of experience the web

- e.g.uva edit quicklist



- taxonomies can be daunting

- personal context for meaning

- some things are hard to bookmark

-session-based

-service-based

-part of page

-wiki vs tagging

-community value

-diverse users

-stimulate local use

-reduce cost of dev

-form community



-using local identities

-post - own data

-add

-recent or popular posts

-rss

-comments

-selected bibs

-active tags



tag based nav

user-based nav

easy url structure - shallow

tags.library.upenn.edu/tag/video_games

-related tags

-combine tags



-items referenced through url server resolver

-then to web



-easy to use -->bookmarklets. Firefox toolbar

-captures title of page, offers structure, recent tags

-privacy options

-rating thing- not much used

-annotation - note vs review

-ability to type kind of post  with *genre*

-  link to groups through title of bibliography

(share across groups?)

- update posts



-requires some local software installation - barrier to use

-context - description

-relatedness

-communities of learners-exploit in class

-title and url are req - url is editable



- 3 table app

- modular output - beginning to support more



-sortable output-by tag, person

-no search of annotations - only tags

-instructions on tagging? Suggestions

-copy urls-assess value in use of copy function



-deployed sept

-positive feedback



whats next?

bundles of stuff

copy others tags

copy based on others tags

order re-order

smart capture of resource md

weighting of resources based on use? Value? Scale? Reputation?

Registration

-multiple tag spaces for multiple uses

-sub.communities

-taxonomy development based on tag analysis

uses

-biliographies

-research guides

-image collection

questions

.learning support

.taxonony

.community

.useful object creation?

.roles [teacher/student]

.annotations and their value

.rights?

Follow up:

Tags.library.upenn.edu/winkler4/cni

washington post tag cloud

netvibes.com

writely.com