Studying the Influence of Scientific Articles on Social Texts Using Word Similarity Analysis and Altmetrics in the Field of Climate Change



The present survey aims to study the scientific texts and social texts linked each other by citations in order to investigate the influence of scientific terms on social texts. The social text is expected to be a bridge between scholars and users by citing scientific texts. By indexing scientific and social text and measuring its similarity by means of an inclusion index, it is possible to see the influence of scientific terms on social texts. The basis of the statistical data of the present study is the scientific papers on climate change issues published in the journals covered by the Web of Science, which have the role of the research community. Sample members consisted of 9876 scientific and social texts (7912 social texts cited 1964 scientific texts) that were extracted and stored in .txt files with the help of Python web scraping using beautiful soup package. The pdf files were extracted manually and converted to a txt file. The indexed terms of the scientific and social texts and the common terms between them, were identified and categorized according to their length and document types separately using the automatic negative indexing model and regular expressions. Measuring the penetration of scientific terms on social texts using the inclusion index, showed that the average penetration rate was 27%. There was a positive, significant and strong correlation (0.76) between the number of common terms and academic penetration. The results of chi-square test showed that there is a significant relationship at 1% level between the term length and the average penetration rate. There is also a difference between the level of penetration and the type of social documents type, including news, blogs, policy documents, Wikipedia, and peer reviews. The highest percentage of penetration is related to policy documents and then Wikipedia. Obtaining 22 citations and 40 social citations on average per paper, and the strong correlation between citation and altmetric score, indicates the citation influence among the academic and non-academic community. Collectively, with the power of altmetrics like policy documents, Wikipedia and blogs and their permeability, they can be used to further promote scientific findings. Publishing scientific findings in these sources, even in a different language from scientific authors, can contributed to greater awareness of society and the general public.