7.1 About Unstructured Text

Data mining algorithms act on data that is numerical or categorical. Numerical data is ordered. It is stored in columns that have a numeric data type, such as NUMBER or FLOAT. Categorical data is identified by category or classification. It is stored in columns that have a character data type, such as VARCHAR2 or CHAR.

Unstructured text data is neither numerical nor categorical. Unstructured text includes items such as web pages, document libraries, Power Point presentations, product specifications, emails, comment fields in reports, and call center notes. It has been said that unstructured text accounts for more than three quarters of all enterprise data. Extracting meaningful information from unstructured text can be critical to the success of a business.