Work in the areas of scientific data visualization, information visualization, and user interface design all contributes to an understanding of how to present document spaces. However, presenting a document space for navigation demands some specialized requirements. Visualization tools address discovery of new information or relations not seen in raw data. For navigation, presentation will address the following issues: specific information to identify places, aids for the path selection process, and help in understanding the structure of a space. The navigation tools are one specific type of user interface.
To present a document space, a document and its attributes are mapped to a visual object and its attributes for display. There are many ways to do this mapping. Many encoding methods have been developed. This section addresses the issue of presentation in three sections. First, the limitations of both humans and display devices are discussed. Next, options for the user interface are outlined. The document space presentation possibilities will be discussed including the problem of mapping data to presentation and of interaction perspectives. Finally, an example of a document space presentation will be shown.
4.1Limitations of the Human Visual System and of Computer Displays 4.1.1Human Visual System Limitations
To design a presentation, it is necessary to realize the limitations of human perception. Many studies have been conducted. Interface design guidelines have been developed based on usability studies of interfaces. One summary of visual information processing in a cognitive perspective related to Web page design can be found in Marks & Dulaney (1997). It includes treatments of sensory constraints, attention, object recognition, and the reading process.
In this paper, consider visual discrimination and pre-attentive stimuli as examples of the human limitations and capabilities. These issues are important in presenting information to a user and are a part of the cognitive process of navigation.
Visual discrimination capabilities as summarized by Spring and Jennings (1993), show the limit of visual perception capability in term of discriminable categories and data type of each visual dimension as shown in Table 3. The discriminable value is computed by dividing the operational range by the discrimination difference. Note that location is computed based on an operation range of 120 degrees in the vertical direction and by 135 degrees in the horizontal. However, a general guideline given by Grether and Baker (1972), cited by Proctor & Zandt (1994), suggests that the number of code steps vs. visual code methods shown in Table 4. The code step, number of distinctive encoded value, indicates lower numbers than the discrimination value. The number of code steps that may be discriminated will depend on the task -- absolute judgment versus relative discrimination of values.
Table 3: Summary of discriminable categories vs. visual sense of objects dimension.
Visual sense of objects dimensions
|
Data type
|
Discriminable categories
|
Nominal
|
Ordinal
|
Interval
|
Ratio
|
Hue
|
x
|
x
|
-
|
-
|
156
|
Saturation
|
-
|
x
|
x
|
-
|
15 – 25
|
Brightness
|
-
|
-
|
x
|
x
|
60
|
Location
|
-
|
-
|
x
|
x
|
15,600
|
Shape
|
x
|
-
|
-
|
-
|
70
|
Size
|
-
|
-
|
x
|
x
|
100
|
Opacity
|
-
|
-
|
x
|
-
|
-
|
Texture
|
x
|
x
|
-
|
-
|
-
|
Table 4: Comparisons of Coding Methods
Code
|
Number of code step
|
Evaluation
|
Maximum
|
Recommended
|
Color
|
|
|
|
Lights
|
10
|
3
|
Good
|
Surface
|
50
|
9
|
Good
|
Shapes
|
|
|
|
Numerals and letters
|
Unlimited
|
|
Fair
|
Geometric
|
15
|
5
|
Fair
|
Pictorial
|
30
|
10
|
Good
|
Magnitude
|
|
|
|
Area
|
6
|
3
|
Fair
|
Length
|
6
|
3
|
Fair
|
Brightness
|
4
|
2
|
Poor
|
Visual number
|
6
|
4
|
Fair
|
Frequency
|
4
|
2
|
Poor
|
Stereo depth
|
4
|
2
|
Poor
|
Angle of inclination
|
24
|
12
|
Good
|
Compound codes
|
Unlimited
|
|
Good
|
Treisman (1986) proposed a model of the visual process in which features, i.e. color orientation, size, and stereo distance, are extracted first, and in parallel -- that is, at the “preattentive” level. While individual features “pop out” in search tasks, combinations of the features do not. The searching time is not dependent on the number of distractors. The second level in the model is a map of locations obtained by means of discontinuation of intensity or color and spatial location by an attention spotlight as a selected focus point. An object is detected based on this information as well as on time and place; identification includes object identity, its properties and relations. This information interacts with a recognition network to identify the object’s “name.”
Preattentive features are not only suitable in locating item of data, but are also used in presenting data value. Healey, Booth, and Enns (1996) show that in a numerical estimation task, preattentive features can be interpreted accurately and rapid. The authors discuss other preattentive features and provide a summary.
In navigation, information from the visual senses will be processed and a decision will be made to select a path. This process will involve a memory of place, of visited places, interpretation of visual coding, etc. It also involves the learning process in using an application interface.
4.1.2Computer Display System Limitations
Computer displays vary widely from small, low-density handheld devices, to wall-size, high resolution displays using projectors. The specification of a screen is given by its viewing area, resolution, color depth, and refresh rate. An ordinary monitor is approximately 13 to 21 inches diagonally across a screen with the intent that it be viewed at a distance of approximately eighteen inches. The height to width ratio is 3:4. Some special monitors have different ratios; for example, some document processing monitors use a 4:3 ratio to mimic paper orientation. Personal Digital Assistant (PDA) devices have small displays limited by pocket and hand sizes. Head-mounted devices have small display areas but are closer to the eye, providing more angles of view than does a standard monitor.
At the current level of technology, computer screens are able to provide more detail than the human eye can distinguish over normal viewing distances. The bottleneck in computer systems is providing fast respond when interacting with high-density graphical objects. An improvement in computer architecture is under way. We will soon see more highly interactive graphical applications and faster response time. However, cost may be limitation. The bigger the screens, the wider viewing angle, the more information we can display at the same time, but the cost of such monitors is a limitation of their deployment.
Share with your friends: |