How are topics born? Understanding the research dynamics preceding the emergence of new areas

Abstract

8The ability to recognise new research trends early is strategic for many stakeholders, such as 9 academics, institutional funding bodies, academic publishers and companies. While the state 10 of the art presents several works on the identification of novel research topics, detecting the 11 emergence of a new research area at a very early stage, i.e., when the area has not been even 12 explicitly labelled and is associated with very few publications, is still an open challenge. 13This limitation hinders the ability of the aforementioned stakeholders to timely react to the 14 emergence of new areas in the research landscape. In this paper, we address this issue by 15 hypothesising the existence of an embryonic stage for research topics and by suggesting that 16 topics in this phase can actually be detected by analysing diachronically the co-occurrence 17 graph of already established topics. To confirm our hypothesis, we performed a study of the 18 dynamics preceding the creation of novel topics. This analysis showed that the emergence of 19 new topics is actually anticipated by a significant increase of the pace of collaboration and 20 density in the co-occurrence graphs of related research areas. These findings are very relevant 21 to a number of research communities and stakeholders. Firstly, they confirm the existence of 22 an embryonic phase in the development of research topics and suggest that it might be 23 possible to perform very early detection of research topics by taking into account the 24 aforementioned dynamics. Secondly, they bring new empirical evidence to related theories in 25Philosophy of Science. Finally, they suggest that significant new topics tend to emerge in an 26 environment in which previously less interconnected research areas start cross-fertilising. 27 28

Scholarly data mining: A systematic review of its applications

“…For instance, Tao et al (2017) have proposed a study map oriented method called RIDP (Reference Injection based Double‐Damping Page Rank) that guides researchers to dig into the underlying principles of a specific paper. However, (Kim et al, 2018; Salatino, Osborne, & Motta, 2017) have studied the keywords associated with each paper in order to analyze the dynamics of research topics and visualize information on the growth and change in focus of research fields. The academic network has been studied by (Tang et al, 2008) to extract and mine academic social networks.…”

“…Social network analysis has been widely used in scholarly data mining applications including citation analysis (Rossetto et al, 2018; Shi et al, 2015; Zhao & Ye, 2013), literature analysis (Dunne et al, 2012; Li et al, 2006; Osborne et al, 2013; Tan et al, 2016; Tang et al, 2008; Tao et al, 2017), document analysis (Kim et al, 2018; Salatino et al, 2017), conference analysis (Effendy et al, 2014), and trend analysis (Hoonlor et al, 2013; Zhang & Guan, 2017).…”

“…Different sub‐areas of computer science have been studied such as artificial intelligence (Alam & Ismail, 2017; Dridi et al, 2019a, 2019b), computational linguistics (Anderson et al, 2012; Asooja et al, 2016; Bakarov et al, 2018; Hall et al, 2008; Paul & Girju, 2009), and big data (Kaempf et al, 2015). Besides, different scholarly data applications have been explored within the area of computer & information science such as document analysis (Alam & Ismail, 2017; Anderson et al, 2012; Salatino et al, 2017, 2018; Bakarov et al, 2018; Paul & Girju, 2009; Tuarob et al, 2016; Caragea et al, 2015), citation analysis (Taskin & Al, 2018; Weller et al, 2011), literature analysis (Dunne et al, 2012; Li et al, 2006; Osborne et al, 2013), conference analysis (Effendy et al, 2014; Effendy & Yap, 2016; Nuzzolese, Gentile, Presutti, & Gangemi, 2016; Tao et al, 2017), and trend analysis (Asooja et al, 2016; Dey et al, 2017; Dridi et al, 2019a, 2019b; Effendy & Yap, 2017; Kaempf et al, 2015).…”