Data Scraping to the rescue: Settling Brazilian public communication of science’s debate?
Author: Marcelo Pereira – UFMG, Brazil
Co-authors:
- Silvania Nascimento – UFMG, Brazil
- Roberto Takata – UFMG, Brazil
To compare the public communication of science outputs in 4 Brazilian public universities: UFMG, UFOP, Unicamp, and UFSCar; we conducted an exploratory analysis of data obtained from public platforms with information about academic outreach activities in those institutions.
Using data scraping techniques, we collated keywords used by researchers in their publications and the description of science outreach activities and projects generating a word frequency table and word cloud. Word lists will be analyzed by ecological techniques (such as Jaccard index) for comparison of biodiversity composition between different areas.
Our hypothesis is that meaning is given by context, which can be revealed by the occurrence and frequency of keywords. If the words that occur in conjunction with the key phrases are similar between the different terms, there is no significant semantic difference between the terms. That analysis had a potential to reveal the semantic dispute in the field: different expressions are used to refer to activities related to public communication of science such as science popularization, public understanding of science, science literacy, among others, highlighting the differences in philosophical point of view between research groups regarding the topic associated with the usage of those terms.
Comparisons between output database platforms will be made too. In the case of UFMG, two platforms are used: one mostly concerned with academic publications, and another specialized in academic outreach activities.
The initial results suggest strong similarities in the keywords related to science popularisation in articles published by researchers of the Universities analyzed. We conducted a preliminary investigation, expecting that that association could be a result of a close collaboration between the institutions, but the data hint otherwise: only one paper in the subject co-authored by members of both institutions was found so far.
Presentation type: Visual presentation
Theme: Technology