Unstructured Data

Unstructured data is stored in its native format with no model of how that data should be interpreted. An example of unstructured data is a novel, there are no set number of words, pages, chapters of plot lines that can be present; you must read the book to know what it contains.

Unstructured data is difficult to analyse as only a qualitative assessment of each item can be made (a book can be good or long or exciting etc.). By contrast structured data allows for quantitative analysis, one number can be directly compared against another with a limited set of outcomes (it is greater than, less than or equal to the other number).

Within human resources it is typical to encounter unstructured data when encountering an applicants C.V. or completing a free text appraisal.

Whilst structured data has significant logical merit this does not mean that unstructured data is undesirable. A mixture of these types of data is common. When designing database for human resource management an important consideration is understanding what types of analyse will need to be performed in the future. This may require bridging the gap between unstructured a structured data by providing additional fields; an example would be to allow the user to record a score for C.V. in the system as well as the document itself. In this way high scoring applications could be searched for whilst the detailed information about them is not lost.