The Venus de Milo was once a block of solid marble until an ancient Greek sculptor negated some of the block leaving what is now one of the most famous statues in the world. Without the tools to remove the unwanted marble, Venus would still be hidden inside the solid block.
In the document analysis world, once BR has clustered documents with its visual classification technology, BR’s negation technology lets users suppress things they don’t want to see so they can focus on what they want to see. Negation empowers users to act on common sense observations about the documents they are analyzing, e.g., “What I’m really interested in is what people entered on the form, not the form itself.” Here are four examples of how negation can be used to improve the analysis of documents within visually-similar clusters.Forms. When dealing forms, once documents are clustered visually, the glyphs (i.e., graphical elements that make up the document) that always appear in a cluster are associated with the “blank” form, i.e., the form prior to people filling in the requested information. When the blank portions of completed forms are negated, the remaining glyphs represent the information filled in on the form. Isolating this information makes it much easier to extract those attributes and analyze that information.
Watermarks and Backgrounds. Negation is useful when dealing with documents containing watermarks or other recurring backgrounds that make it difficult to read or analyze the contents of the document. For example, birth certificates may be printed on paper with special backgrounds intended to make it more difficult for people to falsify the information on them. Such watermarks can make it difficult to convert the glyphs on the page to text, but BR can turn off the background, making for far greater success in extracting accurate text from the documents.
Maps or Engineering Documents. Negation could be used to identify which items were added to an original map or engineering document. Glyphs that were in all versions of the documents would form the “bottom” layer and additional elements from other maps could be represented in layers on top of that layer.
Contract Review. Negation can be used when reviewing collections of contracts to ensure that deviations from standard provisions are being tracked. Negation can operate on a paragraph level, so that once a contract paragraph has been reviewed, it can be turned off on the remaining contracts to be reviewed. The process would continue with approval or review of specific clauses that had not already been reviewed or approved.
The Key: BR’s GGC
Negation and visual clustering is made possible by BR’s global glyph catalog (“GGC”). It catalogs each individual graphical element or glyph on every page of every document, and then uses the GGC to represent the documents to users – when it displays a “document” to a user, it can present only those glyphs that haven’t been negated. Negation can be used at the level of individual glyphs, words, sentences, paragraphs, pages, or documents.
We are just beginning to explore the possible uses of negation for information governance. If you would like to explore how your organization could use negation, please contact Joe Howie at firstname.lastname@example.org.