MiSC: Mixed Strategies Crowdsourcing

Authors: Ching Yun Ko, Rui Lin, Shu Li, Ngai Wong

IJCAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In this work, we propose Mi SC (Mixed Strategies Crowdsourcing), a versatile framework integrating arbitrary conventional crowdsourcing and tensor completion techniques. In particular, we propose a novel iterative Tucker label aggregation algorithm that outperforms state-of-the-art methods in extensive experiments. ... Numerical experiments comparing the proposed Mi SC (mixed strategies crowdsourcing) with pure label aggregation methods are given in Section 5.
Researcher Affiliation Academia 1The University of Hong Kong, Hong Kong 2Nanjing University, Nanjing 210023, China {cyko, linrui, nwong}@eee.hku.hk, lis@smail.nju.edu.cn
Pseudocode Yes Algorithm 1 Truncated higher-order singular value decomposition (SVD), Algorithm 2 Higher-order orthogonal iteration, Algorithm 3 Mixed Strategies Crowdsourcing (Mi SC)
Open Source Code No The paper provides links for
Open Datasets Yes In this section, the proposed mixed complete-aggregate strategies crowdsourcing algorithms are compared with conventional label aggregation methods on six popular datasets, including Web dataset [Zhou et al., 2012], BM dataset [Mozafari et al., 2014], RTE dataset [Snow et al., 2008], Dog dataset [Deng et al., 2009; Zhou et al., 2012], Temp dataset [Snow et al., 2008], and Bluebirds dataset [Welinder et al., 2010].
Dataset Splits No The paper mentions the datasets and their
Hardware Specification No The paper does not provide any specific details about the hardware used for running the experiments (e.g., GPU models, CPU types, memory).
Software Dependencies No The paper mentions
Experiment Setup No The paper describes the algorithms and their mathematical foundations, but it does not specify concrete experimental setup details such as hyperparameter values (e.g., learning rates, batch sizes, number of iterations/epochs), optimizer settings, or other training configurations.