Refining raw data

The amount of work done after the database is created needs to include several other steps. In actuality, textual ETL only produces a raw content file. In order to prepare the raw content for analytical processing, a data refinement step is needed. Fig 12.2 shows this step.

There are several types of refinement that can be made. Some of the refinement processes include sentiment analysis, red flag analysis, summarization, and politeness analysis. After the refinement of the raw data content is done, then the data is ready for visualization.