Real bird dataset with imprecise and uncertain values

TL;DR

This paper introduces real bird datasets collected via crowdsourcing, containing naturally uncertain and imprecise data, to facilitate testing belief function-based data fusion methods.

Contribution

The paper provides publicly available real bird datasets with inherent uncertainty and imprecision, created through crowdsourcing, and discusses how to derive belief functions from this data.

Findings

Datasets contain naturally uncertain and imprecise human contributions.

Method for deriving belief functions from crowdsourced data.

Facilitates testing of belief function-based data fusion approaches.

Abstract

The theory of belief functions allows the fusion of imperfect data from different sources. Unfortunately, few real, imprecise and uncertain datasets exist to test approaches using belief functions. We have built real birds datasets thanks to the collection of numerous human contributions that we make available to the scientific community. The interest of our datasets is that they are made of human contributions, thus the information is therefore naturally uncertain and imprecise. These imperfections are given directly by the persons. This article presents the data and their collection through crowdsourcing and how to obtain belief functions from the data.