Files in this item

FilesDescriptionFormat

application/pdf

application/pdfZHANG-THESIS-2021.pdf (561kB)Restricted to U of Illinois
(no description provided)PDF

Description

Title:Estimation of intrinsic parameters in crowdsourcing problems
Author(s):Zhang, Yichi
Advisor(s):Shomorony, Ilan
Department / Program:Electrical & Computer Eng
Discipline:Electrical & Computer Engr
Degree Granting Institution:University of Illinois at Urbana-Champaign
Degree:M.S.
Genre:Thesis
Subject(s):Crowdsourcing
Binary rating
Vector estimation
Linear minimum mean square error estimation
SVD
Abstract:The term "crowdsourcing" was first coined by Jeff Howe in 2006 to refer to the idea of outsourcing a task to the public. More generally, crowdsourcing means employing the services of a large number of individuals, either paid or unpaid, to acquire information or input into a work or project, often via the internet. In contrast to outsourcing, crowdsourcing frequently involves a broader, less-specific population. Many online services such as Netflix and Amazon, which have strong and efficient recommendation systems, are all using crowdsourcing algorithms to utilize the user provided data to understand the quality of their items. As of 2021, crowdsourcing generally entails leveraging the internet to attract and divide work among people in order to get a cumulative output. Nowadays, a wide range of research projects and applications utilize crowdsourcing and enjoy the benefits including low cost, high speed, high quality, and high flexibility. The widely used Wikipedia is the most successful product that is developed by the use of crowdsourcing techniques. In this thesis, one typical crowdsourcing problem is introduced, and the goal of the problem is to find the unknown intrinsic parameters that stand for the quality of different items. By modeling the problem through a random matrix of observations/ratings, several different algorithms are presented, all of which give the estimations of the unknown vector parameters. Finally, comparisons are made under different metrics, and the conclusion is derived.
Issue Date:2021-12-08
Type:Thesis
URI:http://hdl.handle.net/2142/114014
Rights Information:Copyright 2021 Yichi Zhang
Date Available in IDEALS:2022-04-29
Date Deposited:2021-12


This item appears in the following Collection(s)

Item Statistics