Classifying the content of social media images to support cultural ecosystem service assessments using deep learning models

Cardoso, Ana Sofia; Renna, Francesco; Moreno-Llorca, Ricardo; Alcaraz‐Segura, Domingo; Tabik, Siham; Ladle, Richard J.; Vaz, Ana Sofia

doi:10.1016/j.ecoser.2022.101410

Cited by 28 publications

(14 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The performances of the CNNs adopted to automatically classify the images used in this study align with the ones obtained by the previous applications of transfer learning to the study of CES (Cardoso et al, 2022;Gosal and Ziv, 2020). The performances however were not homogenous for all the classes.…”

Section: Outcomes and Performancessupporting

confidence: 77%

“…Transfer learning allows researchers to adapt freely available pretrained CNNs, such as the ones trained on the ImageNet (Deng et al, 2009) and Places365 databases (Zhou et al, 2017), to new classification tasks. Recently, (Cardoso et al, 2022;Lingua et al, 2022a) have used transfer-learning to create CNNs purposely designed for characterizing the CES provided by natural parks located in the Iberian Peninsula (Spain and Portugal) and British Columbia (Canada) respectively. The use of transfer learning appears to be a promising approach to obtain quick, inexpensive, and detailed data that could be useful for forest CES management.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Assessing forest recreational potential from social media data and remote sensing technologies data

Lingua

Coops

Griess

2023

Ecological Indicators

View full text Add to dashboard Cite

Section: Outcomes and Performancessupporting

confidence: 77%

Section: Introductionmentioning

confidence: 99%

Assessing forest recreational potential from social media data and remote sensing technologies data

Lingua

Coops

Griess

2023

Ecological Indicators

View full text Add to dashboard Cite

“…Previous research has categorised images on social media related to human–species interactions using deep learning (Cardoso et al, 2022; Edwards et al, 2021; Lee et al, 2022). However, our study represents one of the first studies to do this at scale, using crowdsourced data and with spatial metrics linked to specific species, a key research frontier in the field of CES research (Gould, Bremer, et al, 2020), especially in relation to biodiversity (Echeverri et al, 2020).…”

Section: Discussionmentioning

confidence: 99%

“…At the same time, the detail with which particular aspects of biodiversity can be identified using deep learning, including specific species of flora and fauna, means that better connections can be made between biodiversity and CES (Echeverri et al, 2020). For example, Cardoso et al (2022) categorised CES related to biodiversity by identifying images of species on social media using deep learning and a training dataset labelled by the authors themselves. These techniques also allow predictions of very specific ecological characteristics, such as the class of species (Jarić et al, 2020).…”

Section: Introductionmentioning

confidence: 99%

Social media and deep learning reveal specific cultural preferences for biodiversity

et al. 2023

View full text Add to dashboard Cite

Social media has created new opportunities to map cultural ecosystem services (CES) related to biodiversity at large scales. However, using these novel data to understand people's preferences in relation to these CES remains a challenge. To address this, we trained a deep learning model to capture people's interactions with selected flora and fauna on Flickr as a cultural service related to biodiversity and compared this with citizen science data on iNaturalist, with photos of individual species considered as human–species interactions. After mapping the distribution of people's interactions in Great Britain on Flickr and iNaturalist, we find significant spatial differences in people's preferences on the two platforms. Using a second, pretrained deep learning model, we were also able to identify different preferences for species groups such as birds on social media versus citizen science. To better understand people's preferences, we also compared people's interactions with species richness and abundance for a group of 36 bird species, sometimes finding large differences between people's interactions and these ecological measures. Our findings demonstrate that social media can be used to include a wider range of preferences in CES assessments along‐side citizen science data. However, these preferences reflect only a limited first‐hand experience of biodiversity. Read the free Plain Language Summary for this article on the Journal blog.

show abstract

“…This may generate a multitude of intangible and non-market benefits (such as social cohesion), which in turn can hold or give it different values (such as morality, religion, aesthetics, etc.) [ 8 , 9 ]. Schröter et al (2014) conducted a spatial analysis of Telemark County in southern Norway using a spatial model and proposed nine indicators of ecological services, including moose hunting, sheep grazing, logging, carbon sequestration and forest sequestration, avalanche prevention, residential facilities, leisure excursions and the presence of undisturbed areas [ 10 ].…”

Section: Introductionmentioning

confidence: 99%

Connecting Recreational Service to Visitor’s Well-Being: A Case Study in Qianjiangyuan National Park

Zhou

Cai

Chen

2022

IJERPH

View full text Add to dashboard Cite

Natural ecosystems provide service value to people in the region; especially in some nature reserves and national parks, the ecosystem recreational service function is more significant. It is worth paying attention to whether the recreational service function of this ecosystem can serve tourists outside the region and improve their well-being. Taking Qianjiangyuan National Park in China as the research site, based on the Spatially Explicit Ecosystem Services Comprehensive Assessment Model (ESTIMAP), we used the ecosystem services mapping tool to propose a comprehensive assessment framework for ecosystem recreational services and to explore the relationship between recreational services and the well-being of tourists. As a result, we identified the spatial distribution of the supply and demand flow paths of ecosystem recreational services and estimated that the actual flow of ecosystem recreational services was about 12.1 persons/km2 per year. In terms of personal well-being, ecosystem recreational services have a significant positive impact on tourists’ well-being. The service value is concentrated in amusement places and recreational activities, and dynamic recreational activities have a greater impact on tourists. The research results suggest that to improve the well-being of tourists and the value of ecosystem recreational services, national parks and nature reserves should improve accessibility and the participation of tourists in recreational activities.

show abstract

Classifying the content of social media images to support cultural ecosystem service assessments using deep learning models

Cited by 28 publications

References 50 publications

Assessing forest recreational potential from social media data and remote sensing technologies data

Assessing forest recreational potential from social media data and remote sensing technologies data

Social media and deep learning reveal specific cultural preferences for biodiversity

Connecting Recreational Service to Visitor’s Well-Being: A Case Study in Qianjiangyuan National Park

Contact Info

Product

Resources

About