Towards a better understanding of Facebook Altmetrics in LIS field: assessing the characteristics of involved paper, user and post

Facebook mentions to scholarly papers have provided a novel way for reflecting and measuring the process of informal scientific communication. To uncover the underlying mechanism of Facebook Altmetrics, it is essential to investigate characteristics of its contextual data. Take library and information science papers for empirical study, three categories of contextual data were gathered, namely data of mentioned LIS papers, data of Facebook users and data of Facebook post. Hybrid methods including statistical analysis, content analysis and visualization analysis were adopted to analyze the data. Results show that: (1) Positive open access status and active Facebook account would help get scholarly paper mentioned but would not boost the number of Facebook mentions. Number of citations, number of collaborative institutions, and number of collaborative countries showed a significantly positive correlation with the number of Facebook mentions. Health information management was identified to be the most mentioned research topic while bibliometrics and scientific evaluation has received on average the highest number of Facebook mentions. (2) Scientific Facebook users that mention LIS papers were widely scattered geographically but dominated by USA, Spain, Germany, Brazil and Australia. Institutional users (89%) and academic users (84%) are prevailing, especially universities (14%), research institutes (12%), libraries (11%), academic associations (9%) and commercial organizations (8%). (3) Most scientific Facebook posts were relatively short, while the language distribution was less skewed than that of scientific tweets. The post content is mostly a combination of text, links, and pictures and with neutral sentiment. Different types of users have demonstrated significantly different style of content and concerned topics. These findings indicate that Facebook mentions to LIS papers mainly reflect the institutional level advocacy and attention, with low level of engagement, and could be influenced by several features including collaborative patterns and research topics.