Animal behavior and animal personality from a non-human perspective: Getting help from the machine

Research output: Contribution to journalReview articlepeer-review


We can now track the position of every fly's leg or immerse a tiny fish inside a virtual world by monitoring its gaze in real time. Yet capturing animals' posture or gaze is not like understanding their behavior. Instead, behaviors are still often interpreted by human observers in an anthropomorphic manner. Even newer tools that automatically classify behaviors rely on human observers for the choice of behaviors. In this perspective, we suggest a roadmap toward a “human-free” interpretation of behavior. We present several recent advances, including our recent work on animal personalities. Personality both underlies behavioral differences among individuals and is consistent over time. A mathematical formulation of this idea has allowed us to measure mouse traits objectively, map behaviors across species (humans included), and explore the biological basis of behavior. Our goal is to enable “machine translation” of raw movement data into intelligible human concepts en route to improving our understanding of animals and people.

Original languageAmerican English
Article number100194
Pages (from-to)100194
Issue number3
StatePublished - 12 Mar 2021


  • DSML 2: Proof-of-Concept: Data science output has been formulated, implemented, and tested for one domain/problem

All Science Journal Classification (ASJC) codes

  • General Decision Sciences


Dive into the research topics of 'Animal behavior and animal personality from a non-human perspective: Getting help from the machine'. Together they form a unique fingerprint.

Cite this