A Machine Learning Tool to Transform Raw Movement Data into Cleaned, Annotated, and Usable Information

By

Jambon, Aurélie
Albertini, Jean-Vitus
Berton, Fréderic
Corny, Julien
Kappeler, Patrick
Voundy, El-Makki
Rovellotti, Olivier

Type

Article

Date of Publication

2019-6-26

Original Publication

Metabarcoding and metagenomics

Volume

3

Pages

1--2

Subjects

behaviour , biologging , electronic tags , localization , Migration , outliers , Tracking

Contributed by

Pensoft Publishers

Language

English

Published by

Bulgaria Pensoft Publishers 2017-

Identifiers

DOI: https://doi.org/10.3897/biss.3.37480

 

Abstract

<br> The term bio-logging is given to ‘the use of miniaturized animal-attached tags for logging and/or relaying data about an animal’s movements, behaviour, physiology, and/or environment’ (Rutz and Hays 2009). It has been spreading over the scientific world through the last ten years and will become even more popular in the next decade thanks to engineering progresses. Indeed, tags tend to miniaturize, and geolocalization technology, as well as network coverage, is getting improved and cheaper. Today, electronic tags are widely used to study species’ movements and migration, resulting in an enormous amount of data (<br> Li et al. 2015<br> , Thums et al. 2018). Currently, platforms like Movebank enable data storage, visualization and management. However, no easy-to-use tool is currently available for conducting a various array of pre-treatment procedures on those data, such as outliers identification, elevation or time management. In that context, we are developing an open source platform based on algorithms and machine learning to automatically return cleaned and annotated data.<br>