ZoneZoom map navigation for smartphones with recursive view gm

ZoneZoom: Map Navigation for Smartphones

with Recursive View Segmentation

Daniel C. Robbins, Edward Cutrell, Raman Sarin, Eric Horvitz

Microsoft Rearch

One Microsoft Way

Redmond, WA 98052 USA

{dcr, cutrell, ramans, horvitz}@

ABSTRACT

ZoneZoom is an input technique that lets urs traver

large information spaces on smartphones. Our technique

ZoneZoom, gments a given view of an information space

into nine sub-gments, each of which is mapped to a key

on the number keypad of the smartphone. This

gmentation can be hand-crafted by the information space

author or dynamically created at run-time. ZoneZoom

supports “spring-loaded” view shifting which allows urs

to easily “glance” at nearby areas and then quickly return to

their current view. Our ZoneZoom technique lets urs gain

an overview and compare information from different parts

of a datat. SmartPhlow is an optimized application for

browsing a map of local-area road traffic conditions.

Categories and Subject Descriptors

H.5.2 [Information Interfaces and Prentation]: Input

Devices and Strategies, Interaction Styles

General Terms

Human Factors, Design

Keywords

ZoneZoom, SmartPhlow, visualization, hand-held devices,

smartphones, spatial cognition, maps, mobile browsing,

Zoomable Ur Interfaces, Visual Interaction

1. INTRODUCTION

Our technique ZoneZoom is geared toward the emerging

a class of promising rearch that us motion and tilt ur is zoomed-in and wants to then zoom back out from

nsors to control navigation [[16]]. the current view, the ur can either press the number key

Fisheye techniques [[7]], which compress non-focud

parts of the information space into the periphery, may not

work well for maps. This is becau urs rely on

geometric relationships to make judgments about travel

time. Other work has explored adding schematic peripheral

cues to indicate important off-screen information [ [1],

[12]].

3. ZONEZOOM

Our intent is to let urs easily, quickly, and predictably u

one hand to navigate around an infinitely detailed 2-

dimensional information space on a smartphone. We have

focud on the navigation of maps in general and the

browsing of a near-real-time road traffic monitoring

application in specific. As with previous work in this area

we address the issues of information space gmentation,

navigation, and context.

Becau there is no direct pointing device on a smartphone

(such as a stylus or mou) there is no way for a ur to

directly specify an exact location on a map. The fine-

grained interaction required by ZUIs in turn requires a high

degree of attention on the part of the ur – and attention is

a scarce resource in mobile situations. With ZoneZoom we

combine the interactions of panning and zooming into one

“cognitive chunk” [[8]].

In the simplest implementation, a given view of a map at a

particular zoom level can be evenly divided into a t

number of sub-ctors or child views. For a smartphone,

the natural number of sub-ctors is nine, since this

instead of quickly tapping, the ur holds a key down for

more than a moment then releas (press-and-hold), the

current view returns to the view from before the key press.

If the ur is currently zoomed out, such that gment cues

are shown for the current view’s child gments, pressing

and holding on the number key will temporarily zoom the

view to the appropriate child view. Releasing that same key

will then return to the parent (zoomed out) view.

This aspect of spring-loaded view shifting is also en in

consumer flight simulators applications [ [13]] where a

spring-loaded hat-switch on a joystick lets the ur

momentarily glance at different views from the airplane’s

virtual cockpit.

This spring-loaded view shifting allows the ur to quickly

glance at other ctions of the datat without losing track

of their preferred center of interest. This glancing also lets a

ur quickly compare information in nearby regions.

3.4. Non-regular Division of Zoom-levels

The aspect ratio of view sub-gments can differ from that

of their parent or sibling views in order to devote more

Location of

initial view

(4) in context

of larger map

View animation

“pull-back” path

SmartPhlow includes the ability to customize the view

gmentation and feature lection.

7. ACKNOWLEDGMENTS

We thank the Washington Department of Transportation

(WSDOT) for providing the real-time traffic data and

members of the internal SmartPhlow email alias for their

uful feedback and suggestions.

Location of desired

view (6) in context of

larger map

8. REFERENCES

[1] Baudisch, P. and Ronholtz, R. Halo: A Technique for

Visualizing Off-Screen Locations. In Proc. CHI’03, pp. 481–

488.

[2] Bederson, B.B., Czerwinski, M. & Robertson, G.. A fisheye

calendar interface for PDAs: Providing overviews for small

displays. In The Craft of Information Visualization: Readings

and Reflections, pp. 324–344.

[3] Bederson, B., J. Hollan, K. Perlin, J. Meyer, D. Bacon & G.

Furnas. Pad++: A Zoomable Graphical Sketchpad for Exploring

View zooms out during pan from one sibling view (4) to another (6)

Alternate Interface Physics.

Journal of Visual Languages

and Computing

, (7) 1996. pp. 3 – 31.

Figure 5: Animated View Transitions

[4] Bjork, S., L. Holmquist, et al.. WEST: A Web Browr for Small

Terminals. In Proc.

UIST'99

. pp. 187–196.

5. DISCUSSION AND FUTURE WORK

[5] Buyukkokten, O., Garcia-Molina, H. et al. Power Browr:

SmartPhlow has been relead internally within our

Efficient Web Browsing for PDAs. In Proc. . pp. 430–

CHI‘00

organization. Feedback on an internal email list has been

437.

extremely positive, so much so, that the most common

[6] Chen, Y., W.-Y. Ma, et al. (2003). Detecting Web Page

requests has been to add data for other locales and the

Structure for Adaptive Viewing on Small Form Factor Devices.

ability to input personalized routes. In the future we will

In Proc.

WWW’03

, pp. 225 – 233.

engage in formal ur studies along with analysis of usage

[7] Furnas, G. W. Generalized Fisheye Views. In Proc. CHI’86, pp.

patterns (bad on rver statistics).

16–23.

[8] Hinckley, K., R. Jacob, C. Ware. Input/Output Devices and

Maps exhibit many of the same properties as other

Interaction Techniques. In CRC Computer Science and

information spaces: they are continuous, have a

Engineering Handbook, A. B. Tucker, ed. CRC Press LLC: Boca

heterogeneous distribution of interesting data, posss

Raton, FL. to appear.

[9] Holmquist, L. E. and Ahlberg, C. Flip-zooming: A practical

multiple levels of detail, and usually have a highly variant

Focus+Context Approach to Visualizing Large Information Sets.

visual reprentation [ [10]]. Becau of this, we intend to

In Proc. HCI International ’97, pp. 763–766.

investigate the applicability of the ZoneZoom technique to

[10] Hornb?k, K., Bederson, B. B., & Plaisant, C. Navigation

other kinds of information spaces such as web pages,

Patterns and Usability of Zoomable Ur Interfaces With and

spreadsheets, and image browsing.

Without an Overview.

ACM Transactions on Computer-

Human Interaction

, 9(4):362–389.

As noted in the Visual Cues ction, the overview cues are

[11] Jul, S. & Furnas, G. W. Critical zones in dert fog: aids to