%PDF-1.3
1 0 obj
<<
/Kids [ 4 0 R 5 0 R 6 0 R 7 0 R 8 0 R 9 0 R 10 0 R 11 0 R 12 0 R ]
/Type /Pages
/Count 9
>>
endobj
2 0 obj
<<
/Subject (Neural Information Processing Systems http\072\057\057nips\056cc\057)
/Publisher (Curran Associates)
/Language (en\055US)
/Created (2011)
/Description-Abstract (Biased labelers are a systemic problem in crowdsourcing\054 and a comprehensive toolbox for handling their responses is still being developed\056 A typical crowdsourcing application can be divided into three steps\072 data collection\054 data curation\054 and learning\056 At present these steps are often treated separately\056 We present Bayesian Bias Mitigation for Crowdsourcing \050BBMC\051\054 a Bayesian model to unify all three\056 Most data curation methods account for the \173\134it effects\175 of labeler bias by modeling all labels as coming from a single latent truth\056 Our model captures the \173\134it sources\175 of bias by describing labelers as influenced by shared random effects\056 This approach can account for more complex bias patterns that arise in ambiguous or hard labeling tasks and allows us to merge data curation and learning into a single computation\056 Active learning integrates data collection with learning\054 but is commonly considered infeasible with Gibbs sampling inference\056 We propose a general approximation strategy for Markov chains to efficiently quantify the effect of a perturbation on the stationary distribution and specialize this approach to active learning\056 Experiments show BBMC to outperform many common heuristics\056)
/Producer (Python PDF Library \055 http\072\057\057pybrary\056net\057pyPdf\057)
/Title (Bayesian Bias Mitigation for Crowdsourcing)
/Date (2011)
/Type (Conference Proceedings)
/firstpage (1800)
/Book (Advances in Neural Information Processing Systems 24)
/Description (Paper accepted and presented at the Neural Information Processing Systems Conference \050http\072\057\057nips\056cc\057\051)
/Editors (J\056 Shawe\055Taylor and R\056S\056 Zemel and P\056L\056 Bartlett and F\056 Pereira and K\056Q\056 Weinberger)
/Author (Fabian L\056 Wauthier\054 Michael I\056 Jordan)
/lastpage (1808)
>>
endobj
3 0 obj
<<
/Type /Catalog
/Pages 1 0 R
>>
endobj
4 0 obj
<<
/Contents 13 0 R
/Parent 1 0 R
/Type /Page
/Resources 14 0 R
/MediaBox [ 0 0 612 792 ]
>>
endobj
5 0 obj
<<
/Contents 32 0 R
/Parent 1 0 R
/Type /Page
/Resources 33 0 R
/MediaBox [ 0 0 612 792 ]
>>
endobj
6 0 obj
<<
/Contents 34 0 R
/Parent 1 0 R
/Type /Page
/Resources 35 0 R
/MediaBox [ 0 0 612 792 ]
>>
endobj
7 0 obj
<<
/Contents 68 0 R
/Parent 1 0 R
/Type /Page
/Resources 69 0 R
/MediaBox [ 0 0 612 792 ]
>>
endobj
8 0 obj
<<
/Contents 82 0 R
/Parent 1 0 R
/Type /Page
/Resources 83 0 R
/MediaBox [ 0 0 612 792 ]
>>
endobj
9 0 obj
<<
/Contents 94 0 R
/Parent 1 0 R
/Type /Page
/Resources 95 0 R
/MediaBox [ 0 0 612 792 ]
>>
endobj
10 0 obj
<<
/Contents 100 0 R
/Parent 1 0 R
/Type /Page
/Resources 101 0 R
/MediaBox [ 0 0 612 792 ]
>>
endobj
11 0 obj
<<
/Contents 114 0 R
/Parent 1 0 R
/Type /Page
/Resources 115 0 R
/MediaBox [ 0 0 612 792 ]
>>
endobj
12 0 obj
<<
/Contents 132 0 R
/Parent 1 0 R
/Type /Page
/Resources 133 0 R
/MediaBox [ 0 0 612 792 ]
>>
endobj
13 0 obj
<<
/Length 3123
/Filter /FlateDecode
>>
stream
xڵYK۸ϯБsSƎwdS>@$$!Cˇ_
sˉ@h ~|6M.w!&܄AɊ/xS6wDlEnz9?&Q&&L"?m"fO^ɌכB͛'H56L=f0 V9t}kԽJ_jMk.Q:?tG>+/&Y9n@^IBwє'kr՟iﮯ Ɓ$eU!"M#{d瓊I D.^U)f~^xLT8NZq~43\Q7ncZ}GBcU'5l4T[-X]sILt{
cqYUF j2"KOpA<LFSkYy]wg]AW2烰MdboNnIJqŪ命oŷ6