HR4RS stamp
Irene Vega
SDG 13 'Climate action' is the one that arouses the most interest on Twitter within the field of citizen science. This is the main conclusion of a study carried out within the framework of the CS-Track european project, which has analyzed the conversations on the SDGs, the different topics, the most commented, the most important words, the hashtags and the most important users around citizen science and sustainable development goals on Twitter. In addition, by means of the method developed, the impact of this topic can be monitored over time thanks to the replicability of the study, since the procedure carried out is independent of the data analyzed. “It is necessary to know this information to determine the impact that a certain topic has on society and it is necessary to be able to know directly what is happening on Twitter around this topic,” highlights Fernando Martínez Martínez, a researcher at the Higher Technical School of Computer Engineering (ETSII) and co-author of the study.
In this study we have worked with a set of tweets that contained terminology related to citizen science (citizen science, citsci…) and a second filter has been applied to this data set to obtain the messages related to the SDGs. Next, the scientific team has analyzed the labels (hashtags) most used both in retweets (hashtags that appear in tweets that have been shared or retweeted) and in general (adding to the previous count those hashtags that appear in the publications that have not been retweeted). “Although the previous process allowed us to know the total number of times that a hashtag has been used, it was necessary to analyze how these tags evolved over time in order to detect peaks in their use and find important events such as conferences, for example”, details the researcher.
Statistical methods and machine learning techniques for the analysis of tweets
The research carried out has gone a step further and has not only focused on studying the hashtags, but also the rest of the words that appear in a tweet using the TF-ID statistical calculation method, which allows direct visualization of the most relevant terms in the analyzed publications through a cloud of words. However, in order to understand the conversations and extract data of interest, the technique of Topic Modeling LDA. “Through this technique, it has been possible to create subsets of words that together form a topic that is being talked about. For example, in one of the topics created, the words future, earth, biodiversity, conservation, bethechange, renewableenergy, which tell us that one of the conversations is how the individual use of energy impacts the conservation of our ecosystems”, says Fernando Martínez Martínez.
The next step was to classify the tweets by SDGs through a model of machine-learning (computer algorithm to which data is presented and thanks to a previous process in which it is taught to classify, this algorithm gives categories to our data). The results of this process showed that most of the messages published were related to SDG 13, which deals with climate change.
Finally, the relationships between users were analyzed, taking into account the retweets, noting which users most retweeted are headed by the official ODS Twitter account, while the rest are distributed among professional staff, disseminators, bots and project accounts (@cstrackproject o @cs_sdg2020).
Research of citizen science activities
The European project CS-Track (H2020-SwafS-2019-1, grant agreement 872522) is a consortium made up of universities and international research centers with the aim of investigating the activities carried out in citizen science. Within the project, of which the Research groups are part cyberimaginary y LITE, and the Higher School of Telecommunications Engineering of the Rey Juan Carlos University, multiple sources of information are analyzed, among which are social networks. Twitter is one of the most used social networks to share results and create communication channels in citizen science initiatives. Among the trends on Twitter in recent years, the 2030 Agenda stands out. In 2015, the United Nations established 17 Sustainable Development Goals (SDGs) and 169 associated targets to shape global development policies and actions.
Several objectives of the 2030 Agenda are closely related to the objectives that are intended to be achieved in citizen science initiatives, such as the promotion of participation, partnerships and collaborations, education, sustainable living and global citizenship. Studying the conversation about the 2030 Agenda and the SDGs within the citizen science community can help understand how the goals evolve, their relationship with users, and trends at specific times.
Bibliographic reference
D. Roldán-álvarez, F. Martínez-Martínez, E. Martín and PA Haya, "Understanding Discussions of Citizen Science Around Sustainable Development Goals in Twitter," in IEEE Access, vol. 9, p. 144106-144120, 2021, doi: 10.1109/ACCESS.2021.3122086.