The document discusses challenges related to big data, including:
- Big data is fundamentally networked and relational, with value coming from connections between data points rather than size alone.
- There are concerns about digital divides limiting access to big data, as well as ethical issues around how data is accessed and used.
- Definitions of knowledge and objectivity can be challenging with big data, as data lacks historical context and equivalency between different data sources is unclear.
The Sociology of Nothingness: Challenges of Big Data
1. The Sociology of Nothingness:
Challenges of Big Data
Eugen Glăvan
Research Institute for Quality of Life
Towards the Good Society - European Perspectives
Bucharest, 24-26 October 2013
2. Big Data: the horizon of expectations
Change the instruments, and you will change the entire
social theory that goes with them.
Latour (2009)
Big data:
Fundamentally networked. Example: human-centered computing, software to adapt to
changes (MIT Project Oxygen)
The value come from the structure of information itself.
Not because of its size, but because of its relationality to other data.
Digital divides: the poor big data
Ethics
Approach:
Analytic assumptions
Methodological frameworks
Underlying biases – „technological solutionism” (Morozov, 2013)
Questions: what all this data means, who gets access to it, how it is deployed, and to what ends.
(Boyd and Crawford, 2011)
Towards the Good Society - European Perspectives
Bucharest, 24-26 October 2013
4. Big Data: touch and go
Definitions of Knowledge
Harvesters of information - no historical context that is predictive. (Bollier, 2010)
Data traces: from personal networks to „articulated networks” and „behavioral networks”.
Objectivity and Accuracy
„Friends” maximum size of a person's personal network – no one should have a friend list
greater than 150. (Friendster vs. Dunbar in gossip practices)
Bigger Data ≠ Better Data
Twitter – active vs. listen users
Facebook – public access
Data Equivalency
The networks of social media is not necessarily interchangeable.
Towards the Good Society - European Perspectives
Bucharest, 24-26 October 2013
5. UN Woman Campaign
www.unwomen.org/en/news/stories/2013/10/wo
men-should-ads
A series of ads, developed as a creative
idea for UN Women by Memac Ogilvy &
Mather Dubai, uses genuine Google
searches to reveal the widespread
prevalence of sexism and discrimination
against women. Based on searches dated
9 March, 2013 the ads expose negative
sentiments ranging from stereotyping as
well as outright denial of women’s rights.
7. Big Data: Issue Crawler
Network location and visualization software www.issuecrawler.net/
Characteristics:
Crawlers
Databases
Analysis engine
Visualization module
Utilization:
Allows examination of a website, captures the outlinks and performs analysis to
determine points in common, degree of separation or inter-linking.
Retain the more links to / from site to identify limits of the network.
Limits:
The existence of the link does not say anything about traffic
Lack of contextual information
Influence of large and popular sites (search engines or social media)
Towards the Good Society - European Perspectives
Bucharest, 24-26 October 2013
10. Normalized Google Distance – NGD (Rudi L. Cilibrasi and Paul M.B. Vitanyi): using the
probability of occurrence of a term x to extract the meaning of the words in the world wide
web.
f(x) and f(y) the number of results returned from the search terms x and y
f(x,y) denotes the number of pages that contain both terms,
N is the total number of pages that Google indexes
Measures of Semantic Relatedness - MSR)
Big Data: familly science
Towards the Good Society - European Perspectives
Bucharest, 24-26 October 2013
11. Big Data: science, pseudoscience
Towards the Good Society - European Perspectives
Bucharest, 24-26 October 2013