Use cases for the Chinese Buddhist Historical Social Network Dataset

Marcus Bingenheimer

The evolving datasets offered here are prepared in formats that allow immediate use in open source SNA tools (such as Gephi). They are offered for now in two formats .gephi and .gexf: The Gephi specific .gephi is a view of the lay-outed dataset, where nodes are already aligned vaguely along a horizontal timeline. If you have not used Gephi before, perhaps start with this.
Gexf is a XML graph description standard, which contains the raw data, that can easily be transformed into other graph data standards.
For an intro to Gephi see Ken Cherven Network Graph Analysis and Visualization with Gephi (2015), and Devangana Khokhar Gephi Cookbook (2015).

Case 1: You are interested in person X (e.g. the famous 道宣 (596-667)) and want to know about his social network (according to the sources modeled in this dataset)

  1. Load CB-HSNA_version_YYYY-MM.gephi into Gephi
  2. Get the correct 道宣 Person ID of X from http://authority.dila.edu.tw/person/. (You can also search for the name in Gephi (under Data Laboratory > Search), but because of the many-to-many relationship of persons and names, using IDs is preferable.)
  3. To the right find the tab "Filters", go to Topology > Ego Network. Enter the Person ID = Node ID. E.g. A001519 for 道宣:

  4. Extract this Ego Network (Data Laboratory > Export Table as csv. OR: Overview > Use the Select Tool > Right click > Move to new Workspace > File > Export as gexf).
  5. Layout the gexf and prettify:

  6. This can be a starting point to think about Daoxuan in his time and social interactions. Try a 2 or 3 degree ego network to see friends of friends (of friends) of Daoxuan. References to sources are in the Edge List (Data Laboratory).

Case 2: You are interested in viewing the unfolding of the network in a timeline view.

The simplest way to create a timeline slider is:

  1. In the Data Laboratory: Merge Columns > Choose: birthY and deathY > Merge Strategy: Create time interval > Begin: birthY, End: deathY > Parse numbers (0-2000) (I recommend this for now over parsing as dates)
  2. A time line button appears at the bottom > click to enable
  3. Switching back to Overview an adjustable timeline slider is now available

Below timeslider with the CB-HSNA dataset 2019-02. (Fig.1: Network 0-2000, Fig.2: 500CE-750CE, Fig.3: 1000CE-1250CE, Fig.4: 1500-1750).

Note: For many actors birth and death dates are unknown (= Value 0 in the birthY/deathY columns). (There are no actors who are actually born/died in year 0.) The above strategy results in them manifesting at year 0 only. I.e. one can keep them in view if one only extends the timeslider to the right, while leaving it tethered left. Once the left end of the slider is moved the actors with neither birthY nor deathY will vanish. Actors with only birthY will appear after the slider moves across their birthY, but not end (birthY-2000), actors with only deathY will persist from year 0 to their deathY (0-deathY). Obviously this behavior can be changed by changing the 0 values in the respective columns.



home