Page 44 - Ranger Demo

Figure 3
Data characterisation involves the discovery of what the data consists of and deciding whether the data is usable and the organisation should use the data, Fig 3. This phase indexes and tags features within the data source to ensure correct use by the enterprise functions. Currently, we spend the majority of our working time doing data preparation for each task. However, data preparation should be done at an enterprise-level involving the sanitisation, filtering, restructuring and standardisation of the data to ensure its utility before analysis and modelling begin. For example, it draws together structured, and unstructured data and conduct sifts based on classification. In essence, it turns inefficient ‘big data’ into efficient ‘small data’ ready for the organisation to use. The data ingest cycle should be able to standalone and continually operate irrespective of specific RFIs or analytical needs. Only once you have characterised and prepared your data at an enterprise scale do you build your specific analytics and services into that data. The procedures represent the most significant single change in culture and therefore our people, processes and technology that has to occur to realise progression towards data-centric ways of working.
People AMOD
Having considered culture, we next need to consider our people. One element that has been running within NCGI for a while is Analytic Modernisation or AMOD, Fig 4. It has been considering the upskilling of our workforce to prepare ourselves to drive this change:
Figure 4
42