The network paradigm can be used for various disciplines and needs. Essentially this paradigm requires the researcher to see the world as things (i.e. nodes) and relationships between those things (e.g. ties). The classic examples focused on understanding the social relationships between groups through a network perspective (Travers and Milgram, 1969). In this approach, humans are nodes in the network. The ties in the network can express familiarity, friendship, interaction or some other attribute. Networks as a lens for social interaction have motivated researchers to study teamwork in organisations (Cross et al., 2002), adolescence romantic and sexual relationships (, ), friendships forming on Twitter or other social media platforms (Huberman et al., 2009) or interactions in massive multiplayer games (Shen et al., 2014). Obviously, this list is all but exhaustive. Rather, they demonstrate how diverse themes can be studied as networks even when limiting to humans as nodes. Also, as the examples demonstrate, much of these analyses are fuelled by novel data sources capturing more detailed insights on how humans interact. However, ties between individuals can describe different attributes: the similarity between two people or different kinds of social interactions or relationships (Borgatti et al., 2009). Different conceptualisations of ties can be used to motivate different research questions (Borgatti and Foster, 2003).
Beyond social network analysis, studying humans and their social interactions, networks can be used to model transportation, international trade or even gene research (for an extensive list, see Kivela et al., 2014). These examples serve to illustrate the versatility of the network paradigm. For many social scientists, understanding the relationship between actors - both human and non-human - can be seen as a form of a network (Latour, 1990). Following these lines of thinking, researchers have explored in detail relationships between things and actors (e.g. Maier et al., 2018; Hellsten and Leydesdorff, 2019).
However, the relationships studied can be even more abstract. For example, textual data can be presented as a network. Again, the process for exploring textual data as a network requires identifying nodes and what constitutes a tie between those nodes. For example, Bail (2016) created the network by identifying documents' authors (nodes) and words used in the documents (nodes). The ties connected author-nodes to word-nodes, i.e. showed which authors use what type of vocabulary in the documents. Baumer et al. (2018) instead focused on how closely two words (nodes) appear in the documents. If they were in the same sentence and close by each other, they formed a stronger tie between those two words. These approaches illustrate the importance of imagination with network analysis: Almost everything can be seen as a network!