Search

Group decision-making (GDM) requires individuals to weigh and aggregate individual evaluations. Diverse evaluations within a group usually require a consensus reaching process (CRP). However, CRP does not aim to improve group evaluation accuracy. The wisdom of crowds (WOC) applies statistical aggregations to obtain group predictions, without necessitating a certain degree of group consensus. Current WOC techniques identify individual expertise levels based on real outcomes of events, or individuals' meta-information, which might not be obtainable in GDM problems. This study proposes an approach for GDM problems with diverse evaluations provided by individuals with different expertise levels. It weighs individuals based on their contributions to the group, by estimating the evaluated attributes' real outcomes and updating individual weights within a time series. The approach is verified by real collected data and simulated data, then compared with several similar weighting methods and CRP techniques. The results demonstrate the approach's good performance in large groups, with improved identification of expertise levels in diverse groups, without requiring a large number of time points in the time-series. [ABSTRACT FROM AUTHOR]

Sir Francis Galton (Galton, 1907) conjectured the psychological process of magnitude estimation caused the curious distribution of judgments he observed at Plymouth in 1906. However, after he published Vox Populi, researchers narrowed their attention to the first moment of judgment distributions and its often remarkable alignment with the truth, while it became customary to explain this wisdom of crowds effect using ideas of statistics more than psychology, and without considering possible interactions with other distribution moments. Recently, however, an exploration of the cognitive foundation of judgment distributions was published (Nash, 2014). The study not only formalized a possible link between signal detection, evidence accumulation, and the shape of judgment distributions, but also in so doing, conjectured that magnitude estimation by independent individuals causes a systematic error in the wisdom of crowds indicated by judgment distribution skewness. The present study reports findings from an experiment on magnitude estimation and supports these predictions. The study moreover demonstrates that systematic errors by groups of people can be corrected using information about the judgment distribution these people together form, before errors might cause damage to decision making. In concluding, we revisit Galton’s data from the West of England Fat Stock and Poultry Exhibition in light of what we have discovered. [ABSTRACT FROM AUTHOR]

Cultural consensus theory is a model-based approach for analyzing responses of informants when correct answers are unknown. The model provides aggregate estimates of the latent consensus knowledge at the group level while accounting for heterogeneity in informant competence and item difficulty. We develop a new version of cultural consensus theory for two-dimensional continuous judgments which are obtained when asking informants to locate a set of unknown sites on a geographic map. The new model is fitted using hierarchical Bayesian modeling. A simulation study shows satisfactory parameter recovery for realistic numbers of informants and items. We also assess the accuracy of the aggregate location estimates by comparing the new model against simply computing the unweighted average of the informants' judgments. A simulation study shows that, due to weighing judgments by the inferred competence of the informants, cultural consensus theory provides more accurate location estimates than unweighted averaging. The new model also showed a higher accuracy in an empirical study in which individuals judged the location of 57 European cities on maps. • We propose a model extension for Cultural Consensus Theory to two-dimensional continuous data. • The model estimates location judgments considering informants' competence and item difficulty. • The model can be used to derive estimates for unknown locations. • Judgment aggregation with this model extension can outperform unweighted averaging. [ABSTRACT FROM AUTHOR]

Abstract: The Blogosphere is expanding in an unprecedented speed. A better understanding of the blogosphere can greatly facilitate the development of the Social Web to serve the needs of users, service providers, and advertisers. One important task in this process is clustering blog sites. Although a good number of traditional clustering methods exists, they are not designed to take into account the blogosphere unique characteristics. Clustering blog sites presents new challenges. A prominent feature of the Social Web is that many enthusiastic bloggers voluntarily write, tag, and catalog their posts in order to reach the widest possible audience who will share their thoughts and appreciate their ideas. In the process a new kind of collective wisdom is generated. We propose WisColl by tapping into this collective wisdom when clustering blog sites. In this paper, we study how clustering with collective wisdom can be achieved and compare it with a representative traditional clustering method. We present statistical and visual results, report findings and suggest future work extending to many real-world applications. [Copyright &y& Elsevier]

The definition and purpose of crowdsourcing and its relevance to libraries is discussed with particular reference to the Australian Newspapers service, FamilySearch, Wikipedia, Distributed Proofreaders, Galaxy Zoo and The Guardian MP's Expenses Scandal. These services have harnessed thousands of digital volunteers who transcribe, create, enhance and correct text, images and archives. Known facts about crowdsourcing are presented and helpful tips and strategies for libraries beginning to crowdsource are given. [ABSTRACT FROM AUTHOR]