The following tables contain the data we used for building the network of attendance at Rotary meetings. This  attendance network consists in a two-mode network linking individuals (participants in club meetings) with documents (press articles that report on the meeting). Original data comes from three major English-language newspapers in Shanghai. The method for building the corpus of press articles, for extracting name entities (persons) from this corpus and for visualizing the network in Padagraph is described in the "Tables" section. 

The first table contains the master table we used for building the edge and node lists. We provide two different edge lists for constructing two types of networks: one for a multiplex containing every occurrence of every individual in relation to a given document; one that retains only single pairs of relations, adding a weight corresponding to the number of co-occurrences between a given indvidual and a given document (which indicates the strength of the link). Node attributes mostly include the type of node (individual or event), the date of the document/event and the nature of the event. In addition to the edge and node lists, we also included basic network metrics (degree, betweenness centrality) resulting from our analyses in Cytoscape (table 2).