big_data.theory.txt

                                  ┏━━━━━━━━━━━━━━┓
                                  ┃   BIG_DATA   ┃
                                  ┗━━━━━━━━━━━━━━┛

Diff from BI (phases):
  - Data storage + processing:
      - unstructured data ("Varied"): structuring on the fly and dynamic
      - External data (vs internal)
      - Usually big amount of data ("Volume", "Velocity"). Internet of things will create even more.
  - Data analysis:
     - predictive:
         - not past: reporting (what), analysis (why)
         - not present: monitoring
         - importance of data mining and machine learning: subset of AI, when algorithm improves with more data
           (not FUNC args, FUNC itself)
     - Big amount of data -> importance of algorithms and optimization
  - Data visualization: focus on discovery not dashboard

Technologies (usually and as a consequence):
  - Reject SQL (noSQL) and RDBMS
  - MapReduce / Hadoop
  - In memory databases
  - SSD, DRAM
  - Cloud computing

Big data is not real time, but by bringing speed of noSQL databases to Hadoop, tries to become.