Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital...
Transcript of Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital...
![Page 1: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/1.jpg)
Eric Horvitz
Joint work with John Krumm and Adam Sadilek
![Page 2: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/2.jpg)
Behavioral
data store
Population activity in the wild
Paradigm for Digital Experimentation
World
simulator
![Page 3: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/3.jpg)
Behavioral
data store
Population activity in the wild
World
simulator
What-if studies
Paradigm for Digital Experimentation
![Page 4: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/4.jpg)
Behavioral
data store
Population activity in the wild
World
simulator
What-if studies
Paradigm for Digital Experimentation
![Page 5: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/5.jpg)
Behavioral
data store
Population activity in the wild
World
simulator
Studies of incentives & responses
What-if studies
Paradigm for Digital Experimentation
![Page 6: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/6.jpg)
Behavioral
data store
Population activity in the wild
World
simulator
Studies of incentives & responses
Realized design
What-if studies
Paradigm for Digital Experimentation
![Page 7: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/7.jpg)
Decomposition Feasibility Estimation
Human
intellect
Machine
intellect
Availability
Competence
Cost
Interpreter Tasks
Solver
Planner Solution
E. Kamar, S. Hacker, E. Horvitz. Combining Human and Machine Intelligence in Large-scale Crowdsourcing, 2012.
![Page 8: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/8.jpg)
Coordination
(AI, Human)
HUMAN
HUMAN HUMAN
HUMAN
AI
AI
AI AI
AI
![Page 9: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/9.jpg)
HUMAN HUMAN
HUMAN
HUMAN
AI AI AI
HUMAN
HUMAN
AI
HUMAN
Coordination
(AI, Human)
![Page 10: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/10.jpg)
REPAIR
(i)
REPAIR
(i)
TRANS
(j,k)
MT MT MT
TRANS
(i,j)
TRANS
(i,j)
Example: Lingua Mechanica
D. Shahaf, & E. Horvitz. Generalized Task Markets, AAAI 2010.
Search &
Optimization
TRANS
(k,l)
![Page 11: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/11.jpg)
D. Shahaf, & E. Horvitz. Generalized Task Markets, AAAI 2010.
![Page 12: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/12.jpg)
D. Shahaf, & E. Horvitz. Generalized Task Markets, AAAI 2010.
![Page 13: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/13.jpg)
D. Shahaf, & E. Horvitz. Generalized Task Markets, AAAI 2010.
![Page 14: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/14.jpg)
$
Interpreter Tasks
Solver
Planner Solution Perception Action
Human
intellect
Execution Coordination Composition
Machine
intellect
Sensors
Effectors
D. Shahaf, & E. Horvitz. Generalized Task Markets, AAAI 2010.
People Cameras Cranes Forklifts Cars ….etc.
![Page 15: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/15.jpg)
Capture real-world data on location and mobility
Digital experimentation via parameter sweeps
Seek insights on shaping “fabric” of collaboration
Preferences, incentives and links to collaborative fabric
Implications & directions
A. Sadilek, J. Krumm, E. Horvitz. Crowdphysics: Planned & Opportunistic Crowdsourcing for Physical Tasks.
![Page 16: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/16.jpg)
Actively “shaped” routing graph: set proximity d, dwell time t
Canonical Crowd Physics Problem
A. Sadilek, J. Krumm, E. Horvitz. Crowdphysics: Planned & Opportunistic Crowdsourcing for Physical Tasks.
Studies
Locations via GPS-coded tweets
Sweep through d,t
d
Shaped Contact Graphs
t
![Page 17: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/17.jpg)
d
$ d
Contact Graphs from Tweets
A. Sadilek, J. Krumm, E. Horvitz. Crowdphysics: Planned & Opportunistic Crowdsourcing for Physical Tasks.
Actively “shaped” routing graph: set proximity d, dwell time t
Shaped Contact Graphs
![Page 18: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/18.jpg)
d
$ d d ’
d ’
Contact Graphs from Tweets
A. Sadilek, J. Krumm, E. Horvitz. Crowdphysics: Planned & Opportunistic Crowdsourcing for Physical Tasks.
Actively “shaped” routing graph: set proximity d, dwell time t
Shaped Contact Graphs
![Page 19: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/19.jpg)
d =100 meters t =30 minutes
21 days Cities: 60x60 km
Contact Graphs from Tweets
A. Sadilek, J. Krumm, E. Horvitz. Crowdphysics: Planned & Opportunistic Crowdsourcing for Physical Tasks.
Shaped contact graphs: Experiments with design
Actively “shaped” routing graph: set proximity d, dwell time t
![Page 20: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/20.jpg)
Canonical Task: Routing Physical Objects
Synchronization & sequencing of effort
Local vs. centralized plans
Canonical task: Route packages
• Coverage & reachability
• Efficiency (delivery time)
• Optimality of routing heuristics
• Robustness to graph ablation
• Locale sensitivity
![Page 21: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/21.jpg)
![Page 22: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/22.jpg)
![Page 23: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/23.jpg)
![Page 24: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/24.jpg)
![Page 25: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/25.jpg)
![Page 26: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/26.jpg)
Random infinite graphs
Static, homogeneous networks, repeat lattice structure
Studies: Properties of finite, real-world contact graphs
Leverage for task coordination
![Page 27: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/27.jpg)
Lévy flight: random walk w/
heavy-tailed probability distribution.
Frequent short-hops, rare long jumps (Mandelbrot)
P(x) a x-b b = 2.2
Small-world phenomenon?
Graph diameter O(log n)
Shortest path discoverable?
![Page 28: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/28.jpg)
Global: Dijkstra, Floyd-Warshall
Assume: Complete, static knowledge
Real-time: Uncertainty about future locations
Local opportunistic routing
Use heuristics based on historical data
![Page 29: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/29.jpg)
60 x 60 km bounding box around Seattle, 6 months
450 x 450 m cells, start & destination (ci, cj )
Test: Final 35 hours
Train: Rest of 6 mos.
Local policy:
> For each location l, consider statistics of proximity to l. > Keep package or handoff based on historical proximity
Global: Dijkstra
![Page 30: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/30.jpg)
Phase transition @
d =350 meters t =30 minutes
![Page 31: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/31.jpg)
![Page 32: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/32.jpg)
![Page 33: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/33.jpg)
![Page 34: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/34.jpg)
![Page 35: Eric Horvitz - microsoft.com€¦ · Capture real-world data on location and mobility Digital experimentation via parameter sweeps Seek insights on shaping “fabric” of collaboration](https://reader033.fdocuments.net/reader033/viewer/2022043017/5f39b72b7de39236ed4b7dc7/html5/thumbnails/35.jpg)
Understanding & Shaping Collaborative Systems
Studies of crowd physics for sensing & acting in world
Design, optimization of collaborative substrate
Rich area, multiple directions for enhancing collaboration
Opportunities include new approaches to epidemiology,
e.g., where reachability & efficiency are minimized