4.4.4 Structured and unstructured data

• Structured data


  Information held in databases with clearly defined attributes whose types are known in advance
• Unstructured data


  Data that has either not been structured in advance, or is unsuited to structuring
• Unstructured data applications


  • Identifying a house or person from a photograph
  • Finding stories in a digitised newspaper archive
• Sources of unstructured data


  • Machine-acquired data (satellite/aerial photography, scientific data, sensor data)
  • Human-generated content (business documents, emails, social media)
• In the last few years, the demand for unstructured data has begun to outstrip that for structured data
• A 2014 study estimated that unstructured data applications used 69 EB of the 80 EB of the data storage shipped that year, and the proportion of storage and processing devoted to unstructured data continues to increase
• Uses of big data relevant to everyday lives


  • Targeted advertising in online services
  • Analysing patient mobile phone use to identify health changes
  • Finding links between genetic abnormalities and health conditions
• The topic of personal data inevitably causes controversy because we all have different feelings about what is and is not acceptable