Logo Lanfrica
  • Home
  • Atlas
  • Insights
  • Docs
  • Sign in

© 2026 Lanfrica. All rights reserved. All copyrights of the resources shown on the Lanfrica website belong to the original copyright holders, unless explicitly stated otherwise.

Data Sheet 1_Large language models enable large-scale analysis of human-bird relationships in South African cities.docx

Domain:

environment and energynatural language processing

Record type:

paper
Creator:
SagChe
Host:avatar

Human perceptions of urban wildlife can shape conservation priorities and public support for biodiversity initiatives, however research on human-bird relationships remains spatially biased towards the Global North. Here we assessed the perceptions of 36 urban bird species across four South African urban contexts using a mixed-methods approach. Survey respondents (n = 1,977) rated species likeability on a 5-point Likert scale and provided open-ended explanations for their ratings. Quantitatively, South African urban birds were generally well-liked, with notable variation among species: the Malachite Kingfisher (Corythornis cristatus, mean ± SE = 4.91 ± 0.02) and Orange-breasted Sunbird (Anthobaphes violacea, 4.91 ± 0.02) scored highest, and the Common Myna (Acridotheres tristis, 2.50 ± 0.03) scored lowest. To analyse the approximately 71,000 open-ended responses, we employed ChatGPT, a generative AI large language model, to identify eight themes underlying species appeal. The highest-rated species were primarily valued for aesthetic appeal and emotional connections, while the lowest-rated species were associated with aggressive behaviours and negative ecological impacts. Factor analysis revealed three perceptual clusters, demonstrating that some species evoke multidimensional responses whilst others are viewed through a single dominant lens. Notably, aesthetic patterns did not universally predict appeal and many highly rated raptor species were valued for emotional connections rather than physical traits. Additionally, negative perceptions did not apply uniformly to all non-native or problematic species, with some receiving moderately positive responses despite ecological concerns. These findings highlight the complexity of human-bird relationships in urban contexts and demonstrate that large language models can enable qualitative analysis at large scales. By offering an African perspective, this study contributes to a more inclusive understanding of how urban residents perceive and value birds.

Visit

figshare.com

Tags

Zoologyartificial intelligence (AI)ChatGPTcitizen scienceGlobal Southhuman-wildlife relationshipsurban biodiversityurbanisation

Licenses

CC BY 4.0

Similar

Can Small-Scale Data Poisoning Exacerbate Dialect-Linked Biases in Large Language Models?Romanization-based Large-scale Adaptation of Multilingual Language ModelsTharuChat: Bootstrapping Large Language Models for a Low-Resource Language via Synthetic Data and Human ValidationMitigating Data Scarcity for Large Language ModelsComparative Analysis of Large Language Models in Moroccan Darija TranslationLarge Language Models for Sentiment Analysis to Detect Social Challenges: A Use Case with South African Languages

Can Small-Scale Data Poisoning Exacerbate Dialect-Linked Biases in Large Language Models?

Style-conditioned data poisoning is identified as a covert vector for amplifying sociolinguistic bia

Romanization-based Large-scale Adaptation of Multilingual Language Models

Large multilingual pretrained language models (mPLMs) have become the de facto state of the art for

TharuChat: Bootstrapping Large Language Models for a Low-Resource Language via Synthetic Data and Human Validation

The rapid proliferation of Large Language Models (LLMs) has created a profound digital divide, effec

Mitigating Data Scarcity for Large Language Models

In recent years, pretrained neural language models (PNLMs) have taken the field of natural language

Comparative Analysis of Large Language Models in Moroccan Darija Translation

Large Language Models for Sentiment Analysis to Detect Social Challenges: A Use Case with South African Languages

Sentiment analysis can aid in understanding people's opinions and emotions on social issues. In mult