Notes
Slide Show
Outline
1
Multi-threaded Interactive Storytelling
  • Kim Pearson
  • Ursula Wolz
  • Program in Interactive Multimedia,
  • The College of New Jersey
2
The Question:
  • How to support a journalist in thinking differently about multiple story lines
  • Why? A web site is problematic for story line
    • The Nancybelle project has multiple angles
    • A web site loses the sense of a narrative arc
    • The Emergent Behavior of the pathways can thwart story impact
3
Who Are We?
  • Kim:  A journalist concerned with finding new ways to engage audience with rich, complex non-fiction narratives.
  • Ursula:  A computer scientist (former computational linguist) interested in deep structures of multimedia information dissemination.
4
What Are We Trying To Do?
  • Long term:
    • Kim: (the journalist): Design a writing system for creating news stories for database driven, object-oriented narratives.
    • Ursula (the comp scientist): understand deep structure of multiple story lines
  • Short term:
    • Kim: Abstract the process for doing literary journalism
    • Ursula:  how can we go beyond the relational database to support this
5
Collaboration with CS
  • Computing is not merely in service of journalism
  • Nor is computing driving journalism
  • It must be a partnership
  • An example from The College of New Jersey
    • content management for an online news magazine
    • http://www.tcnj.edu/~unbound  (in collaboration with Monisha Pulimood, CS)
6
Problem in Journalism:
  • Via emerging media, how can we engage audiences that have turned away from news
  • Trying to embrace and understand online & interactive media
    • Emphasis on:
      • social media, citizen journalism
      • making news content accessible on multiple platforms
      • genre - presentation tools, content management systems, games
7
Emerging Critique from CS
  • Technology should allow users to extract news that matters to them.
  • Push vs. pull systems
    • Delivery versus choice
    • When does the story get in the way
    • When does the story get lost because of the complexity of the information (data)
8
Examples of Critique
  • Holovaty: Creates relational databases for news and information:
      • http://ChicagoCrime.org
      • Post Remix: http://blog.washingtonpost.com/postremix/
    • Holovaty sacrifices story – but arguably, losing story loses the essence of journalism
    • Thomson Financial -- robots write basic earnings stories (knowledge-based agents)
9
Magazine Writer Perspective
  • How to employ this technology in literary journalism
  • What might be gained or lost in the process
  • Chose prototype story of Nancybelle Valentine
    • A rich collection of images
    • Preserve a narrative arc in telling her story
    • Can technology capture the multiple stories?
10
The Initial Vision of a Web
11
Emergent Behavior
  • From the author, not even to the user yet
    • How to organize the links
    • How to define the size of a chunk
    • How to choose paths
  • A Relational Database of Course
    • But how to view it
    • As a chart (a table with links)
12
The Granularity Problem
  • How big is a “chunk” of information and what does it looks like?
    • A text snippet
    • A photo (or worse a montage)
    • A video clip
    • An audio clip
    • A set of “links”
13
The Perspective Problem
  • Implied narrative captured in the pathways through the piece
  • Is the “scene” the right level of granularity?
  • How do you re-use these “scenes”
  • Who controls the story:
    • The writer (the computer supports the organization)
    • The computer (acting as editor, imposing structure on the writer)
    • The user ???  introducing severe emergent behavior
14
The Abstraction Paradox
  • Allows you to construct a larger whole
  • Allows information to be accessed and presented from a variety of perspectives
  • Reduces human labor (simplifies coding)
  • Introduces structure where it may be unwanted
  • Absolutely must get the interface right --- prototype!
15
Unbound as an Example
  • Original site was hard-wired
  • Maintained via extensive human labor
  • Current site is easy to update
  • Getting there required extensive conversations about the editorial process so that the process could be abstracted
16
Nancybelle
  • Takes the complexity to a new level
  • Technique comes from New Journalism:
    • Narrative arc
    • Emphasis on character development
    • Scene-by-scene construction
    • Extensive dialogue
    • Experiment with point-of-view
    • Symbolic detail
17
The Hero Narrative
18
Representation of Symbolic Detail
19
 
20
The Emerging Database
21
More than Relational
  • Relational DB only defines the relation between scenes
22
More than Relational
  • Relational DB only defines the relation between scenes
  • Narrative Arc imposes an overarching structure
23
More than Relational
  • Relational DB only defines the relation between scenes
  • Narrative Arc imposes an overarching structure
  • To identify that structure we need to experiment with real stories such as Nancybelle
24
Turning the Tables
  • Is there a place for multi-threaded narrative?
  • Do users need
    • story
    • narrative arc
    • structure?
  • Data mining, data synthesis, where’s the story?