environmental tagging
DESCRIPTION
a short presentation about Environmental taggingTRANSCRIPT
Tagging in the Real WorldStudy of sustainability-related issues
Nicolas Maisonneuve
WP2: SONY CSL ContributionDelivrables 2.4, 2.5
Outline
NoiseTube.net (3rd Year)
Zexe.net (2nd year)
Ikoru: Armin Linke’s Installation (during the 3 years
Tagging usage in theartistic community
Tagging usage for sustainability- related issues
Tagging usage in the real world
Social
Location (GeoTagging)
Social
Tagging the user experience (in the real world)
Location (GeoTagging)
Social
Sustainability
Pollution exposureSocial justice Carbon Footprint…
Tagging the user experience (in the real world)
Social Justice: Zexe.net (Eugenio Tisseli)
2008 - Campaign in Geneva about the life of handicapped people
Zexe.net = a community memory for representing daily experiences using Folksonomies (via pictures and sound files)
Several campaigns for un(der)-represented communities (Taxi drivers Mexico, Disabled people Geneva, Motoboys Brazil)
Tagging « slices of life ».
Collective Level - Adaptive sensor network at a low cost- Living map showing the shared experience to noise
Green user experience- Phone = environmental instrument- Autonomy to measure noise pollution
Noise Pollution: NoiseTube.netNoiseTube Participatory approach to monitor noise pollution using mobile phones
- Raising awareness (extension of zexe.net principles)- Scientific issue: lack of real data
Issue 1: Hazard identification
Only measurements, No semantic information
Simulated mapMeasurement done by real sensors
New tagging usage: Use people as semantic sensors
Issue 1: Hazard identification
Only measurements, No semantic information
Simulated mapMeasurement done by real sensors
Issue 1: Hazard identification
Contextual Tag cloud
Searching by value = Hard for non-experts Example: meaning of 75 dB(A) ? , lat,lng={2.34,12.5} ?
Issue 2: Searching/navigating in a large dataset of environmental data
Geographical space
Numerical space
Searching by value = Hard for non-experts
Issue 2: Searching/navigating in a large dataset of environmental data
Geographical spaceSemantic space
Numerical space
Semantic exploration of measurementsvia rich context
Limitation of social tagging (not enough data) Enriching the context via automatic generation of contextual tags
Automatic generation of contextual Tags
Social tagging
Roadwork Neighbors
Automatic generating of contextual Tags
Social tagging
Roadwork Neighbors
Machine Tagging = set of classifiers Example : Loudness Classifier
<50 dB “Quiet”
[50, 75] “Annoying”
>85 dB “risky”
[75, 85] “noisy”
Automatic generating of contextual Tags
Social tagging
Roadwork Neighbors
“High variation”
Loudness Signal Pattern
“short-term risky exposure”
Automatic generating of contextual Tags
Social tagging
Roadwork Neighbors
Loudness Signal Pattern
Location
Location type
“outdoor” (with gps)
Street name: “rue Amyot” (Google Map API)
Type: “indoor”
Street name
City Name: “Paris”
City Name
Automatic generating of contextual Tags
Social tagging
Roadwork Neighbors
Loudness Signal Pattern
Time Week: “working day” , “weekend”
Day: “Morning” , “afternoon”, “evening”,”night”
Season (+ GPS sensor): “summer”, “spring”
LocationDay
Week Season
Automatic generation of contextual Tags
Social tagging
Roadwork Neighbors
Loudness Signal Pattern
Weather Conditions Winds: “calm”, breeze” , “storm”
Temperature: “freezing” , “fair”, “hot”
type: “Cloudy”, “raining”,etc..
LocationTime
(At the city level)
Temperature:
TemperatureWinds
type
User-generated tags
Roadwork Neighbors
Loudness Signal Pattern
LocationTime
Weather
Machine-generated tags
Automatic generation of contextual Tags
Semantic profile of the context
Semantic exploration
Automatic generation of contextual Tags
Participatory monitoring of noise pollution using mobile phones
Demo