How Topics Affect Twitter Attention

The purpose of the investigation conducted was to discover trends in twitter popularity regarding different areas of science. This investigation can benefit areas of marketing such as targeted advertising, as well as demographic research in order to correctly test certain demographics and obtain research grants. Results included possible confirmation of our motive through principal component analysis, The data was compiled using RStudio and was narrowed down by subjects, Altmetric scores, and countries. The data was parsed through to find Key words in the abstracts of articles. Principal Component Analysis was applied to a matrix of padded tweet dates, arranged by subject. These arranged dates were also plotted to visualize trends over time. From the data collected, the articles that were most tweeted about, between January 1st, 2016 to July 1st, 2016, worldwide were articles concerning physics. Out of all the articles, ”death” was the keyword most popular in articles’ abstracts. Disease-related words appeared far more often than the word ”cure”. The United States of America, Canada and Great Britain had the highest number of tweeters. Great Britain’s population was mainly interested in articles regarding dentistry, while Canada and the United States of America had a higher tweet count in articles related to health science.

本研究旨在探究不同科学领域在推特（Twitter）上的传播热度趋势。本研究可应用于精准广告等营销领域，亦可服务于人口统计学研究，助力特定人群特征的验证与科研基金的获取。研究结果可通过主成分分析(Principal Component Analysis)验证初始研究动机；数据集依托RStudio进行整合，并按研究主题、Altmetric评分与国家维度进行筛选。研究对数据集进行解析，以提取学术论文摘要中的关键词。以研究主题分组的填充推文日期矩阵被用于主成分分析，同时对分组后的日期数据进行可视化绘图，以展示随时间变化的传播趋势。在2016年1月1日至2016年7月1日期间的全球数据中，被转发提及最多的学术主题为物理学领域的论文。在所有论文的摘要关键词中，"死亡"为最热门的关键词；与疾病相关的词汇出现频率远高于"治愈"一词。美国、加拿大与英国的推文发布者数量最多，其中英国用户主要关注牙科学相关论文，而加拿大与美国用户对健康科学领域论文的转发量更高。