Link Search Menu Expand Document
Uni Notes

Lecture Notes

CSE 442

Table of contents
  1. Week 2 Tuesday: Data and Image Models
  2. Week 4 Tuesaday – Interactive Design

Week 2 Tuesday: Data and Image Models

  • Task, data, and domain \(\to\) processing algorithms and mapping \(\to\) image

Data Models / Conceptual Models

  • Data models are formal descriptions
  • Conceptual models are mental constructions
  • Schneiderman 1996: one-dimensional (sets and sequences), temporal, two-dimensional (maps), three-dimensional (shapes), n-dimensional (relational), trees (hierarchies), networks (graphs)
  • Nominal, ordinal, quantiative (interval, ratio)
  • Dimensions: discrete, usually independent, variables describing data
  • Measures: dependent variables, data values that can be aggregated (usually quantiative)

Relational Data Model

  • Represent data as a table or relation
  • Relational algebra (Codd 1970) – operations on data tables, table in table out
  • Roll-up (data along desired dimensions) and drill-down

Tidy Data (Wickham 2014)

  • Every variable forms a column
  • Every observation forms a row
  • Every type of observational unit forms a table
  • ‘Normalized forms’in database theory

Common Data Formats

  • CSV
  • JSON

Image Models

  • Jacques Bertin – The Semiology of Graphics, using imagery to encode information, the first theoretical work in visual encoding
  • Visual language is a sign system
  • “Resemblance, order, and proportional are the three signfields in graphics.”
  • Visual encoding variables: position, size, value, texture, color, orientation, shape, tlransparency, blur/focus, length/area/volume

Formalizing Design

  • Assuming \(k\) visual encodings and \(n\) data attributes
  • We want to pick the best encoding among \((n+1)^k\) possible encodings
  • Principle of consistency, principle of importance ordering
  • Design criteria (Mackinlay 1986): expressiveness and effectiveness
  • “tell the truth and nothing but the truth”
  • Effectiveness of encodings by data type
  • Mackinlay’s design algorithm, 1986 “A Presentation Tool”; user formally specifies data model and type, tests expressiveness, and generates encodings that pass test

Week 4 Tuesaday – Interactive Design

  • Tukey, PRIM-9, data exploration – discovery in multiple dimensions
  • Brushing and linking across views
  • Interaction techniques can allow us to make input/output relationships to discover new insights
  • Direct manipulation: visual representation of objects and action, rapid and reversible actions, selection by pointing / not typing
  • Grammar of graphics