Transcript Wireless Networking with Android

Emerging Ubiquitous Knowledge
Services: From Mobile Sensing to
Ubiquitous Crowdsourcing and Beyond
Uichin Lee, Howon Lee*, Bang Chul Jung**, Junehwa Song***
KAIST Knowledge Service Engineering
*KAIST Institute ICT
**Kyungsang University
***KAIST Computer Science
Oct. 11, 2011
Knowledge Service Innovation
• Traditional knowledge services: mainly delivered
by experts with domain specific knowledge
• Dramatic change of knowledge services due to
recent advances of ICTs (information and
communication technologies) and networked
collaboration among people
Smart Devices
Web 2.0
Social
Networking
Major Contributors
•
•
•
•
Human: crowdsourcing, human computation
Device: processing, sensing, networking
Network: in-network services
Application: Web 2.0, service mashup, data mining
Human
Device
Content
Fixed
access
Crowdsourcing
Human computation
Smart home/office
Network
Internet
Content provider
Application
Radio
access
On the move
Networking
Applications
Our focus: ubiquitous knowledge services with
mobile sensing and crowdsourcing
Mobile Sensing
• Sensors in smart devices
(smartphones, pads):
accelerometer, magnetometer,
gyroscope, light, proximity,
camera, voice, GPS
• Wireless communications:
Bluetooth, Wi-Fi, 3/4G
• Sensor applications:
– User experience enhancement
• Resizing screen/tilt,
• Gaming, augmented reality
– Mobile sensing apps:
• Traffic information systems
• Location-based services
Crowdsourcing
• “Crowd" + “Outsourcing"
– “distributed labor networks over
the Internet to exploit the spare
processing power of millions of
human brains” – best example,
Wikipedia
• Related concepts: collaborative
system, wisdom of crowd, human
computation, collective
intelligence, wikinomics
인간계산
노동
위키피디아
소셜 Q&A (지식인) 마켓플레이스
인소싱(insourcing)
아웃소싱(Outsourcing)
크라우드소싱(Crowdsourcing)
해결책의 폭
모바일 소셜네트워크
Ubiquitous Knowledge Services
스마트폰을 이용한
모바일센싱
Ubiquitous Knowledge Services
• Goals:
– Seamlessly integrate content from various sources at
large scales
• Content: data, information, knowledge
• Sources: databases, grassroots (sensors, humans)
– Infrastructure dependence is minimal (compared to existing
approaches)
– Also derive new values for end users in ways that the
contributor of the content did not plan or imagine
Examples of Ubiquitous Knowledge Services
교통정보수집
공동체의식제고
도시계획
현장조사
모바일 UX 테스트
연구지원
범죄수사
Contents
• UKS Applications
Potholes
– Vehicular apps
• Traffic engineering, ride quality
monitoring (cracks, potholes)
– Community-awareness apps
(e.g., health and wellness)
– Social sensing with Twitter
• UKS Platform Design
– Unique features
– UKS platform
Air pollution (CO2)
Social networks
Pothole Patrol
• Acceleration data gathering from vehicles (geo-tagged)
• Simple data processing to detect a pothole, and statistical
processing (clustering) for accurate detection
Smooth Road
Pothole
The Pothole Patrol: Using a Mobile Sensor Network for Road Surface Monitoring, Eriksson et al, MobiSys, 2008
Community Awareness:
Health and Wellness
• Personal environmental impact report (PIER) on
“health and wellness”
• Participants use mobile phones to gather location
data and web services to aggregate and interpret
the assembled information (e.g., air pollution, CO2
emission, fast food exposure)
"Sensing Pollution without Pollution-Sensors”
Existing
Infrastructure
Annotation
/Inferences
Scientific
Models
Activity Classification
e.g., staying, walking, driving
Air pollution CO2
exposure emissions
(PM 2.5)
Fast food
exposure
Tracklog format
GIS Data Annotation
e.g. weather, traffic
Weather, traffic data
Impact and Exposure
Calculation
School,hospital,fast
food restaurant
locations
User profile
Data Aggregation
PEIR, the Personal Environmental Impact Report as a Platform
for Participatory Sensing Systems Research, Mun et al., Mobisys 2009
Social Sensing with Twitter
Event detection from twitter
detect an
earthquake
search and
classify them into
positive class
some users posts
“earthquake
right now!!”
Object detection in
ubiquitous environments
detect an
earthquake
Probabilistic model
Probabilistic model
values
Classifier
tweets
・・・ ・・・
・・・ ・・・
・・・
observation by twitter users
some earthquake
sensors
responses
positive value
observation by sensors
earthquake
target event occurrence
target object
Earthquake shakes Twitter users: real-time event detection by social sensors, Takeshi et al, WWW 2010
Building Ubiquitous Knowledge
Service (UKS) Platform
• Research community:
– ArchRock, SensorBase (UCLA), SensorMap (Microsoft), IrisNet
(CMU), and many others..
• Standardization efforts:
–
–
–
–
Semantic Web Enablement (SWE)
Semantic Sensor Web (SSW)
Machine-to-Machine (M2M)
Internet of Things (IoT), e.g, EPCglobal Network
• Essential features of UKS:
– Humans are part of the systems (e.g., crowdsourcing, human
computation)
• Exploiting social networks is also important
– Due to resource constraints (e.g., battery), mobile operations
must be optimized, which may require service-device/-network
interactions
– System should perceive the intention of a user and provide the
customized knowledge services at any place and any time
Building Ubiquitous Knowledge
Service (UKS) Platform
• Key components of UKS:
– Smart device, knowledge gateway, knowledge server
– User agents, service-device/network interactions and
optimization, content integration/negotiation
Conclusion
• Knowledge service innovation with recent advances of ICT
(i.e., device, network, and application intelligences) and
networked collaboration among people
• Ubiquitous knowledge services aim at (1) seamlessly
integrating content from various sources at large scales,
and (2) deriving new values for end users in ways that the
contributor of the content did not plan or imagine
– Example services: vehicular applications (pothole detection),
community awareness (health and wellness), SoundSense, event
detection with Tweets
• Building a ubiquitous knowledge service (UKS) platform
that can integrate grassroots knowledge (and also existing
content) and dynamically generate new values to end users