1. 'These developments reveal a diversity of perspectives, interests, and expectations regarding Big Data' (Ekbia, et al., 2014:1524)
2. Authors identify recurring themes and problems
   1. disciplined-specific
   2. issue-oriented
   3. methodologically-informed
   4. Authors' synthesization frames these all as 'dilemmas'
3. Discussion of phenomenon vs. appearance, and the gap between them
   1. 19th Century – save the phenomenon, produce the appearance, via causal mechanisms
      1. 20th Century – save the appearance, let go of causal explanations
         1. 21st Century – predict the appearance, via Big Data
         2. Where phenomenon becomes appearance – and raises the question of whether data, itself, is enough?
4. 'Big Data methodologies ... collapse the distinction between phenomena and appearances altogether, presenting us with structuralism run amok' (2014:1530)
5. Big Data processing
   1. mechanical
   2. human judgment
      1. human labor and expertise
   3. 'human-machine symbiosis' (2014:1535)
6. quantitative vs. qualitative
   1. 'traditional tension ... is rendered obsolete' with Big Data (2014:1531)
7. '[d]ata need to be imagined and enunciated to exist as data, and such imagination happens in the particulars of disciplinary orientations, methodologies, and evolving practices' (2014:1531)
8. issues encountered with Big Data analysis
   1. 'cleaning'
   2. sharing/transparency
   3. privacy
      1. reidentification
   4. 'what counts and what doesn't'
      1. sampling
      2. selection bias
      3. relates to scale and scope of the data being analyzed
         1. historical dillemas remain the same
            1. relevance
            2. validity
            3. generalizability
            4. replicability
   5. open data or not
      1. intellectual property
   6. ethics
   7. creator contribution or exploitation
   8. creator compliance or resistance
   9. 'Big data is dark data ... [need to] devise techniques that bring human judgment and technological prowess to bear in a meaningfully balanced manner' (2014:1539)
   10. political economy specific issues
       1. wealth creation: winners vs. losers
       2. 'filter bubbles' and recommender systems (2014:1538)
9. Big Data doesn't relieve a researcher of responsibility for his/her work
10. Data visualization
    1. accuracy vs. aesthetic appeal
    2. may be weighted toward aesthetic appeal
    3. are tradeoffs noted to end users?
    4. see figure 2 (2014:1533)
    5. '[t]he complexities associated with understanding Big Data and the visual capabilities of the digital medium have pushed this kind of practice to a whole new level, heavily tilting the balance between truth and beauty toward the latter' (2014:1533)
11. Digital divide: haves vs. have nots
12. 'Real or near-real-time information processing – that is, delivery of information and analysis results as they occur – is the goal and a defining characteristic of Big Data analytics' (2014:1534)
13. Big Data attributes
    1. autonomy
    2. opacity
    3. generativity
    4. disparity
    5. futurity
14. 'The duality, futurity, and disparity of Big Data, along with its various conceptualizations among practitioners, make it unlikely for a consensus view to emerge in terms of dealing with the dilemmas introduced here' (2014:1540)
15. Will require a 'concerted effort between academics and policy makers, in rich conversation with the public' (2014:1541)
16. Academic interest in Big Data
    1. Increasing publications on the topic
17. Non-academic interest in Big Data
    1. Big Data offers opportunities for
       1. commerce
       2. innovation
       3. social engineering
       4. Users include business, industry, and government
18. General public interest in Big Data
    1. Fostered by 'the blurring of boundaries between data produced by humans and data about humans' (Ekbia, et al., 2014:1524, c.f. Shilton, 2012)
19. Definition remains undefined, with perspective dependent upon the definer/purpose
    1. product-oriented
       1. focus is quantitative and on the size of data
          1. volume
          2. velocity
          3. variety
          4. value
          5. veracity
          6. This perspective centers on the 'attributes of the data themselves' (2014:1526)
    2. process-oriented
       1. relating to collection, curation, use of data
          1. storage
          2. management
          3. aggregation
          4. searching
          5. analysis
          6. This perspective centers on the 'processing and storing abilities of technologies' (2014:1526)
    3. cognition-oriented
       1. concerns the way humans relate to data
          1. human limits when faced with Big Data
          2. need for mediation, including 'trans-disciplinary work, technological infrastructures, statistical analyses, and visualization techniques' (2014:1527)
          3. This perspective centers on what is needed for humans to make sense of the data
    4. social movement - added by the authors
       1. socioeconomic, cultural, and political shifts
       2. diffusion of technological innovation
       3. complex ecosystem
       4. strategic partnerships
       5. This perspective accounts for gaps 'between these visions and the realities on the ground' (2014:1528)