Data Studies 2021 // S08. Networks and platform-specific data

Pablo Velasco // Information Studies // pablov.me

Plan for the day:

• Graph theory & Sociometrics
• ANT epistemology & networks
• Digital Methods and Digital Methods Initiative
• Digital Objects: hyperlink & hashtag
• TCAT server

The seven bridges of Königsberg problem: is it possible to cross each bridge once and return to the starting point?

Graph theory

Leonhardt Euler solved the problem, and inaugurated the field of graph theory: all points must have an even degree (an even number of connections).

Baran (1964)

• Structures based on Nodes and Edges
• Statistical/Quantitative relations: clusters, centrality, density
• Centralized, Decentralized and Distributed

Sociometrics

• Moreno (1934)

  • Discover social life through network properties & topologies

  • Analysis of relational data

  • Gain insight into social relations and make them available to intervention (Guggenheim 2012) -> participants become observers of their own problems

Gießmann, S. (2017). Jacob Levy Moreno, Sociometry, and the Rise of Network Diagrammatics

ACTIVITY: trace a network of the class

• What are the nodes?
• What are the edges?
• What is the starting point?
• How do you measure the weight of the edges?

Go to: https://kinopio.club/ds21s08-XQXwaV8m0qglvM_h3BElR
(you need to sign up, but not obligatory!)

Actor Network Theory(ANT)

Actor Network Theory emphasizes a "flat" approach to the social, and argues that it can be study as a network consisting of any kind of elements.

Latour, B. (2007). Reassembling the Social: An Introduction to Actor-Network-Theory

ANT

• Relations between “technology” and “society”
• 1-level: neither nature and society, nor micro and macro
• Actors: computers, text, humans, etc (and mostly the places they "collide")
• Properties emerge from networks

Latour, B., Jensen, P., Venturini, T., Grauwin, S., & Boullier, D. (2012). ‘The whole is always smaller than its parts’–a digital test of Gabriel Tardes’ monads.

• No point of departure (but an epistemological point of entry)

More subtle than the notion of system, more historical than the notion of structure, more empirical than the notion of a complexity, the idea of network is the Ariadne's thread of these interwoven stories (Latour 1993)

Actors in a social network:

• By their engagement roles:
  • sources
  • leaders
  • commenters
  • lurkers
  • etc.
• By their human / non-human categorization:
  • bots
  • hashtags
  • servers
  • etc.
• By their structural role:
  • bridges
  • hubs
  • stars
  • etc.

How to do empirical research from this epistemological approach?

Digital methods

Digital Methods (big umbrella)

• Digital methods: “the use of online and digital technologies to collect and analyse research data"
  Snee et al (2016). Digital methods for social science

• Snee et al include:

  • web-based surveys
    Dillman, D. A. (2011). Mail and Internet Surveys: The Tailored Design Method
  • online interviewing and focus groups
    Kazmer, M. M., & Xie, B. (2008). Qualitative Interviewing in Internet Studies: Playing with the media, playing with the method
  • computer mediated discourse-analysis
    Herring, S. C. (2004). Computer-Mediated Discourse Analysis
  • digital ethnographies e.g. virtual ethnography
    Hine, C. (2000). Virtual Ethnography

Redistribution of methods

Marres, N. (2012). The redistribution of methods: On intervention in digital social research, broadly conceived

• redistribution of research: not so much an opposition between IT firms and researchers (e.g. Savage and Burrows 2007), but a reconfiguration of agents in social research

• 4 views on the redistribution of methods (along a spectrum):

  • methods as usual: old social methodologies incorporated into digital devices
  • big-methods: vast datasets allow us to perform large-scale analysis on real network dynamics**
  • virtual methods: adaptation of the social research methods into the digital
  • digital methods: adapt digital devices for the purposes of social research

"Natively digital" methods & objects

The rise of the Internet enables new research methods that deploy specifically digital devices such as links, comments and shares
Rogers, R. (2009). The End of the Virtual: Digital Methods

• methods embedded in online devices: of crawling, scraping, folksonomy
• digital objects: tweet, username, timestamp, hyperlink, hashtag
• built upon existing services

Places of research (spheres)

Rogers, R. (2013). Digital Methods

1. mid to late 90s: hyperlink, individual website analysis
2. early to mid 00s: blogosphere, search engine critique
3. late 00s: location-aware, web 2.0, social media

"cross-spherical analysis”: web, blogs, news, twitter, etc

Digital Methods Initiative

https://wiki.digitalmethods.net/Dmi/ToolDatabase

Digital object: hyperlink

Brin, S., & Page, L. (1998). The anatomy of a large-scale hypertextual Web search engine.

DMI tool: Issuecrawler

https://issuecrawler.net

Digital object: hashtag

Early use: IRC(Internet Relay chat)

Ad-hoc hashtags

• ad hoc : emergent "as necessary" / "when needed"
• coordination between distributed, and possibly disconnected actors
• '#'
  • user generated
  • competing specific themes and uses
  • evolving and non-deterministic
• community: common topics and (direct) engagement
• issue network (Marres): comunities generated from contested topics

Top hashtags in the Climate Change space on Twitter over time (March-Jun 2012)

(see project information here)

Visual Network Analysis

(tbc in workshop)

• evolving, non-deterministic data
• force-vector: gravity
• not spatial, but relative position of the nodes (in relation to each other)
• borders are not exact, communities are fluid and can overlap
• mixed methods: qualitative inquiry before, after, or alongside the quantitative analysis

Venturini, T., & Jacomy, M. (2014). Visual Network Analysis

Based on Venturini et al (2014), which structures can you identify in this network? Is it significant, regardless of its specific content?

DMI tool: TCAT

https://ds.cc.au.dk/
(access code on workshop page)

• API based
• Built by keywords OR geolocations
• Can be thoroughly queried (if you have a complicated query on a big bin, it will take some time)
• Different data and formats can be exported from it (depending on the amount of data and the bin size, this process may take some time)