Premium
Volunteer reputation determination in crowdsourcing projects using latent class analysis
Author(s) -
Moreri Kealeboga
Publication year - 2021
Publication title -
transactions in gis
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.721
H-Index - 63
eISSN - 1467-9671
pISSN - 1361-1682
DOI - 10.1111/tgis.12713
Subject(s) - crowdsourcing , credibility , reputation , latent class model , reliability (semiconductor) , class (philosophy) , source credibility , computer science , data science , knowledge management , artificial intelligence , world wide web , machine learning , political science , power (physics) , physics , quantum mechanics , law
The production of information in crowdsourcing projects from multiple, often unorganized, individuals makes it a challenge to assess the value of volunteers' contributions. Moreover, measures of how the reliability and credibility of this information can be established, based on a contributor’s reputation, remain a challenge. Statistical methods can be used to develop quantitative indicators that model the relationship between volunteers and their contributions. For example, the reputation of volunteers can be determined based on the extent to which their multiple contributions are regarded as reliable by their peers, using a robust latent class analysis (LCA) methodology. This study presents an innovative LCA approach as a suitable technique for establishing the credibility of volunteers in a crowdsourcing initiative. This methodology was successfully applied in a case study in Botswana, which demonstrated its practicality and effectiveness in determining the credibility of volunteers, supplying information for a land administration project in the country.