Retro-labelling is the process of applying protective marking to documents that have already been created. It relies on the documents being properly classified first and, until the DataCube, has been a difficult and time-consuming process.
Retro-labelling is important for organisations that have had no labelling system, and have a large legacy of unstructured, unlabelled documents in their systems. It is also important for organisations that have implemented the current view of the Government Protective Labelling system, with its 6 tiers and must now replace it with the new 3 tier system.
Hundreds of terabytes of electronically stored information, accounting for millions of documents going back decades is not unusual in large organisations such as Local Authorities. The task of reviewing, categorising and labelling of this volume of information would be impossible to do manually. Apperception’s DataCube automatically categorises and labels hundreds of millions of documents with a high degree of accuracy in a fraction of the time it would take to do any other way.
DataCube uses advanced textual analysis techniques, based on latent semantic indexing, to analyse all of the documents. It then compares each document to examples of the kinds of documents which should be categorised in each category. For Local Authorities, we have already complied a detailed list of over 25,000 exemplars relating to the Local Government Classification Scheme (LGCS).
Having categorised each document, DataCube then applies protective marking (labels) to each one. These labels are based on the default protective marking value for each category in the classification schema.