~/talk

Boomerang: Rebounding the Consequences of Reputation Feedback on Crowdsourcing Platforms

on YouTube 20m
Talkat UIST 2016 · 2016-10-16..10-19· [developer-experience, microservices]

speakers

name:Snehalkumar Neil S. Gaikwad
links:none on file
name:Durim Morina
links:none on file
name:Adam Ginzberg
links:none on file
name:Catherine Mullings
links:none on file
name:Shirish Goyal
links:none on file
name:Dilrukshi Gamage
links:none on file
name:Christopher Diemert
links:none on file
name:Mathias Burton
links:none on file
name:Sharon Zhou
links:none on file
name:Mark Whiting
links:none on file
name:Karolina Ziulkoski
links:none on file
name:Alipta Ballav
links:none on file
name:Aaron Gilbee
links:none on file
name:Senadhipathige S. Niranga
links:none on file
name:Vibhor Sehgal
links:none on file
name:Jasmine Lin
links:none on file
name:Leonardy Kristianto
links:none on file
name:Angela Richmond-Fuller
links:none on file
name:Jeff Regino
links:none on file
name:Nalin Chhibber
links:none on file
name:Dinesh Majeti
links:none on file
name:Sachin Sharma
x:ssharma2211
name:Kamila Mananova
links:none on file
name:Dinesh Dhakal
links:none on file
name:William Dai
links:none on file
name:Victoria Purynova
links:none on file
name:Samarth Sandeep
links:none on file
name:letshine Chandrakanthan
links:none on file
name:Tejas Sarma
links:none on file
name:Sekandar Matin
links:none on file
name:Ahmed Nasser
links:none on file
name:Rohit Nistala
links:none on file
name:Alexander Stolzoff
links:none on file
name:Kristy Milland
links:none on file
name:Vinayak Mathur
links:none on file
name:Rajan Vaish
links:none on file
name:Michael S. Bernstein
x:mrb_bk

resources

video:https://youtu.be/c2ez2WZ0jqs
slides:none on file

summary

A reputation system called Boomerang is built into the demo crowdsourcing marketplace and uses three-point ratings (check+, check, check−), cascading task release for requester-prioritized early access, and task-feed ranking for workers. Field experiments with 35 workers show Boomerang reduces rating inflation by ~50%, increases neutral and negative ratings (2.5× and 1.8×), and yields a near 0.9 SD effect toward accuracy; the system therefore rebounds feedback into concrete allocation of early task access.

AI-written summary · about

chapters (8)

  1. 0:00Introduction to the project
  2. 1:09The reputation problem
  3. 2:45Incentive design approach
  4. 6:28Boomerang design
  5. 10:41Experimental evaluation
  6. 12:51Broader implications
  7. 15:36Conclusion and future work
  8. 16:40Q&A session

related talks

~/search
type to search 6,249 conferences, 138,987 speakers and 185,172 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.