Understanding human-nature interactions using social media data

New research shows how social media data can help to understand the versatile human activities in and the preferences towards nature. Such understanding is essential for effective nature conservation.

Understanding human-nature interactions is crucial for finding successful solutions for nature conservation. Versatile information is needed for making better decisions that help address the biodiversity crisis and support the wellbeing of people. Big data, such as social media data are revolutionizing the knowledge generation in conservation science and practice. The information shared online by ordinary people becomes valuable research material through the combination of advanced analytical methods. 

Researchers at the University of Helsinki have been analyzing social media posts for several years already. The aim has been to study the interactions between people and the environment, especially in national parks, but also in other green areas. In a recent article published in journal Biological Conservation, members of the research group bring together their present multidisciplinary understanding on how social media material can be utilized in conservation science. All the authors have contributed equally for the preparation of the article.

– The aim is to share the research team's broad methodological expertise with others. In addition, we want to provide new ideas for utilizing social media as a data source for different information needs, says Associate Professor Tuuli Toivonen, leader of the Digital Geography Lab.  

Spatial information in focus  

From the viewpoint of nature conservation, social media platforms such as Facebook, Twitter, Instagram, Flickr and Chinese Weibo are the most interesting sources of information for researchers. Users can add spatial information to the content trough adding coordinates or mentioning a place name in their update. Spatial information can be used to analyze, for example, the most popular places in a national park, the location of species observations, or the regional reactions of an emotional event. 

– The available information varies by social media platform. Flickr and Instagram primarily contain images and videos of nature and activities, while Twitter updates usually tell about opinions, attitudes or interest towards certain topics. Putting these on a map is extremely interesting, says researcher Vuokko Heikinheimo who is doing a PhD on the use of social media data for conservation science. 

 Artificial intelligence for content mining  

Social media is a treasure trove for researchers. The masses of individual posts offer new dimensions to studying human activities, values and networks across time and space. With artificial intelligence, knowledge extraction becomes even more efficient.  

– Computer vision and language technology enable us, for example, to automatically identify human activities and species in visual and textual content in social media, but there is still a long way to go to placing these observations in context to understand the intentions behind a social media post, says Assistant Professor Tuomo Hiippala, who is familiar with automatic content analysis methods. 

When interpreting the results, one must acknowledge the limitations of current algorithms, especially as social media updates include a wide variety of languages and imagery. It is also important to remember that only a fraction of people share their thoughts and experiences in social media. Hence, the results may be biased to describe the activities of certain types of people only.  

– We know that young people and women are more eager to share nature photos and moods in social media compared to other groups of people. Spontaneously sent messages provide different information than the information collected by surveys or counters, but these sources of information complement each other, explains Tuuli Toivonen.  

Valuable information for decision-making  

Social media provides fast and cost-effective new perspectives on different phenomena along with information from other sources. When using social media data, one must also pay attention to ethical issues, as well as the terms of use of the various services. It is the responsibility of the researcher to protect the privacy of people, even if the data has been shared publicly online. Understanding biases in the data is also a key part of the research.   

– Social media create a deluge of data that can be used in conservation science to understand human nature interactions. Biodiversity conservation is notoriously underfunded. Therefore, we are developing and using methods from machine learning and natural language processing to efficiently investigate illegal activities that may threaten biodiversity conservation in order to inform policy making, says Adjunct professor Enrico Di Minin, who has recently received funding from the European Research Council to study illegal wildlife trade on social media.  

Reference:

Social media data for conservation science: A methodological overview,
Toivonen Tuuli, Heikinheimo Vuokko, Fink Christoph, Hausmann Anna, Hiippala Tuomo, Järv Olle, Tenkanen Henrikki, Di Minin Enrico. https://doi.org/10.1016/j.biocon.2019.01.023

Additional information:  

Associate professor Tuuli Toivonen, Digital Geography Lab
Email: tuuli.toivonen@helsinki.fi   
Twitter: @TuuliToivonen

Doctoral student Vuokko Heikinheimo, Digital Geography Lab
Email: vuokko.heikinheimo@helsinki.fi   
Twitter: @Vuoggis