Apa kabar ? Extracting sociological data from masses of Indonesian newspaper clippings
description
Transcript of Apa kabar ? Extracting sociological data from masses of Indonesian newspaper clippings
Apa kabar? Extracting sociological data from masses of Indonesian newspaper
clippings
What if?In 4th largest nation in the world…
… with its chaotic bureaucracy…
… but its vibrant press…we could extract vital sociological trends automatically from its dozens of digital/ digitised newspapers?
Elite network shifts during regime changeA digital humanities approach to network analysis using Indonesian language
electronic newspaper archives
Regime change
194519651998
Elite shiftsOld regime
Decays internallyCrumbles rapidly
Creative & destructive ferment
New ruling coalition
Sociology and computersHistorical sociologists know a lot alreadyKITLV catalogue
1945 - 457 titles on “Indonesian revolution” 1965 - 643 titles on “PKI” (Indonesian
Communist Party – destroyed after putsch) 1998 - 1042 titles on “Reformasi”
But too many elite actors to follow – trees and forest
Can digital humanities show us global patterns?Mountains of data, mathematical modeling, complexity
NetworkNode (vertex)
IndividualOrganisation
Line (edge)Undirected Directed (polar)
AnalysisPolarity (friend/ foe)Authority (no. incoming)Small world (shortest dist.)Cohesion, centrality,
brokers, cliques…. over timeEtc etc etc!
Normal elite networksCentralised Decentralised
AuthoritarianCohesive national
ruling class, cosmopolitan (weak
ties)
Locally rooted cliques (strong ties), little
rotation
DemocraticPlural national elites/ fragmentation/ intra-
elite competition
Locally rooted cliques, frequent
rotation
Elite network shiftsCentralised Decentralised
Authoritarian 1998 1945
Democratic 1965
Digital humanitiesAutomatically read digital/ digitised Indonesian
newspaper archives for 1945 and 1998KITLV, NIOD, KB, WayBackMachine
Extract names (individuals, organisations, places)
Build elite networksUndirected (related if in same article)Polar (sentiment analysis - friend/ foe)
Analyse networks - actor-centred, context-oriented
Analytical boundariesWho is “elite”? What is an elite “network”?
Positional, decisional, reputational, or relational?
Influence (foxes), or domination (lions)?National, regional, or community?
What do newspapers (not) reveal about elite networks?They “manufacture consent” (highly filtered)
Lots of discussion needed!
InterdisciplinaryKITLVNIODInformatica UvAInformatika ITBDANSErasmus Studio,
Erasmus U Rot
Research objectivesIndonesian PhD – read newspapers,
produce networksPostdoc 1– interrogate networks
sociologicallyPostdoc 2– interrogate networks
mathematicallyProduce demonstrable prototype softwareBasis for future collaboration