a study on mental representations for realistic visualization. the particular case of ski trail...
TRANSCRIPT
A STUDY ON MENTAL REPRESENTATIONS FOR
REALISTIC VISUALIZATION:
THE PARTICULAR CASE OF SKI TRAIL MAPPING
R.Balzarini, A. Dalmasso, M.Murat
Commission II, WG II/6
1
GeoVIS 2015: Rendering and Cognition with Images and
Hybrid Visualizations
source : Balzarini, Buttarazzi, SIG 2014
R.Balzarini GeoVIS15 2
Introduction
a work in progress collaboration between geographers (INRIA-GRA), historians (LARHRA) ,computer and
cognitive scientists (LIG), around the study of the evolving representation of a territorial model:
the design of ski resort trail maps.
Research project :
MECOMO – MEmory, COgnition and MOdelling of mountain landscape
This work has been supported by:
the LABEX Innovation en Territoires de Montagne ANR-10-LABX-50-01
the Research Federation Innovacs, UPMF Grenoble-CNRS
the Inria Equipex Amiqual4Home ANR-11-EQPX-0002.
R.Balzarini GeoVIS15 4
Introduction
Case study: what are ski trails maps?
Are iconic images of the sport of skiing and its relation to the terrain on which ski resorts are built (Tait, 2010).
Panoramic views are oblique perspectives of any angle and they often are not topographically accurate.
source : A. Tait, Cartographic perspectives, 2010
R.Balzarini GeoVIS15 5
Introduction
Case study: what are ski trails maps?
Besides artistic talent, the panoramist must possess the
ability to read 2D topographic source materials and translate
this information into a graphical 3D representation.
Source: Patterson, 2000
Ski areas often ask for the mountain to “look bigger” and for the artist
to distort the mountain for this purpose. (Patterson, 2000).
Sou
rce:
Ate
lier
No
vat,
19
88
R.Balzarini GeoVIS15 6
Overall issue
They are at a turning point : paper ski maps no longer
meet the needs of a large part of the customers.
The question arises of adapting to digital practices.
For over 40 years mountains operators have been
providing a representation of GI through artistic
creation of panoramas and ski maps.
tracing the evolution of GI’s representation within the ski resorts mapping tools
advocating recommendations to mountain operators
R.Balzarini GeoVIS15 7
Methodology
An exploratory approach to answer the main research questions:
1. Are the artist and user’s mental representations the same?
2. What representations make ski maps effective to perform a user’s task?
Ch
ab
ris
et K
oss
lyn
, 20
05
Research questions of the study
To answer to the research questions, an experimental protocol was set up:
1. Analyse of expert-activity when making ski maps
2. Analyse of user-activity when reading ski maps
3. Comparison between the provided and used information, to identify expert-representations
handled by users.
R.Balzarini GeoVIS15 8
Methodology
Source : MECOMO project, 2015
Modalities: AN was interviewed in "ecological conditions" according to thinking-aloud and explanation techniques during 3 hours and 20 minutes.
1. The experimental protocol to study expert-artist activity
Participant: Arthur Novat, CEO of the company Atelier Novat
Material: AN could use color pencils, paper, ski plans archives from “Atelier Novat” and a PC.
Data: A set of "messy data" (sketches, diagrams, videos and audio recordings) was collected (Chi, 2006). Video-audio recording of the interview was transcribed.
R.Balzarini GeoVIS15 9
2. The experimental protocol to study user-skier activity
Methodology
Participants: 20 subjects, between 18 and 65 years old, 12 women and 8 men.
Modalities: “controlled conditions”. Semi-structured interviews to perform tasks (location, way-finding, decision-making), according to thinking-aloud and explanation techniques.
Material: paper ski maps in real formats and felt pens.
Data: about 15 hours of video-audio recording were transcribed, sketches, drawings.
Alpes d’Huez ski trails maps, made by Atelier Novat
Villard ski trails maps, made by Atelier Novat
Subject during the interview
R.Balzarini GeoVIS15 10
Methodology
3. Data analysis
Verbal analysis is a methodology for quantifying the qualitative coding of the contents of verbal
utterances (Chi, 1997).
An excerpt of utterances segmentation from expert’s activity analysis (Arthur Novat’s corpus)
An object describes the temporary grouping of a collection of visual features together
with other links to verbal-propositional information (Ware, 2008).
R.Balzarini GeoVIS15 11
Preliminary results Analysis from verbal data and from sketches led to qualitative results.
1. the identification of rules in expertise procedure and user practice 2. graphic objects organized in taxonomy of conceptual categories 3. information arising from the graphics
Generic rules in practice of ski maps
Mainly rules for making ski plans from Pierre & Arhur Novat activity
R.Balzarini GeoVIS15 12
Analysis from verbal data and from sketches led to qualitative results.
1. the identification of rules in expertise procedure and user practice 2. graphic objects organized in taxonomy of conceptual categories 3. information arising from the graphics
Preliminary results
1. Geography
1.1 Domain boundaries 1.2 Sunlight exposure, orientation (shadows contrast) 1.3 Focal point
2. Geomorphology
2.1 Terrain profile 2.2 Peaks and ridges 2.3 Slopes (stiff, craggy …) 2.4 Corridors 2.5 Hollows, combes 2.6 Rocks, cliffs 2.7 Fir trees 2.8 Snow and ice (hues variance)
3. Tracing
3.1 Colored ski runs 3.2 Ski runs geometry 3.3 « Green » areas 3.4 Ski lift
4. Structures
4.1 Pictograms 4.2 Buildings 4.3 Roads
5.1 Names of the ski runs 5.2 Names of the ski lift 5.3 Elevation values 5.4 Toponyms
5. Nomenclature
R.Balzarini GeoVIS15 13
Preliminary results
Analysis from verbal data and from sketches led to qualitative results.
1. the identification of rules in expertise procedure and user practice 2. graphic objects organized in taxonomy of conceptual categories 3. information arising from the graphics
Comparing
the information the artist-expert intends to convey
and
the information that is interpreted by a skier.
R.Balzarini GeoVIS15 14
Decision making in a discovery approach of a ski area,
is not without hesitancy among skiers.
Discussion on users’ representations and difficulties
Difficulties may be manifested by misunderstanding, uncertainty, inconsistency and illegibility
about representations.
From a total of 55 expressions of difficulties
geomorphology 30.9%
tracing 27.3%
structure 18.2%
geography 16.4%
nomenclature 7.3%
R.Balzarini GeoVIS15 15
Discussion on users’ representations and difficulties
Real position of Auris en Oisans and Alpes d’Huez from Google maps images©2015
Deformed position of Auris en Oisans and Alpes d’Huez from Atelier Novat.
“I prefer not to take care of the real topography but mostly take care of representation as I want to give”.
R.Balzarini GeoVIS15 16
Perspectives
Expert’s tacit knowledge analysis
A quantitative approach on skiers’ use of
graphic objects.
an eye-tracking study
a large-scale online questionnaire.
These results provided an inventory of
expert and users representations and
allow refining future hypothesis.
Sou
rce:
SM
I web
sit
e
R.Balzarini GeoVIS15 17
Conclusion
Grateful thanks to Arthur Novat for the time
and enthusiasm he has devoted to the project.
The cognitive approach brings deep knowledge to the study of
evolution of realistic representation into the digital era.
Ski map is a "hijacked artifact"
born as a support to contemplation and became a support for decision making