Leveraging Structured/Unstructured Data in Enterprise Content Management with 3DI

In the data landscape, documents in shared storage, CMS, and scanned archives are often labeled as "unstructured". However, a closer inspection reveals an inherent structure and interconnectedness within these documents. With the help of 3DI's three core elements - Graphical Analysis, Content Analysis, and Attribute Analysis - we can unlock this structure and derive meaningful insights.

Imagine a matrix with document types running horizontally and document attributes or data elements vertically. Each cell that's checked off indicates a corresponding data element present in a given document type. With hundreds of document types in a typical large enterprise, you start to grasp the complexity and vastness of this matrix.

The title graphic offers a sneak peek into how 3DI intertwines document types and attributes, crafting the information fabric of an organization.

Maximizing the Power of the Matrix with 3DI

The data housed in the document attribute matrix can supercharge content management systems, making them more efficient and enhancing their capabilities by providing fielded data to bolster search functionality and generate reports for select documents.

Furthermore, cells in the matrix can be linked to fields and tables in business process systems. Consider a financial institution, for instance. A database-driven loan processing application in such a setting would likely have a field indicating whether a borrower authorized a credit check. This entry must correspond to a credit check authorization form signed by the borrower. Using 3DI, this form can be accessed via the matrix by looking under the document type "credit check" for either the loan number or the borrower’s social security number.

Unveiling the Matrix with 3DI

Detecting and defining an organization's inherent document attribute matrix requires a practical, data-driven approach. The organization can certainly define document-type labels, organizational structures, and formatting guidelines. However, it must work with the existing data elements—at least in the short term. 3DI's Attribute Analysis is key in this endeavor.

3DI: Categorizing Documents by Type

Graphical Analysis, the first core element of 3DI, is instrumental in identifying document types that populate the matrix. By automatically grouping all documents—native electronic and scanned paper—based on graphical similarity, it simplifies the process immensely.

The number of graphically similar clusters is usually less than 1% of the total documents. The largest clusters house most documents in an organization. Assigning a document type to the entire cluster becomes easy, requiring a review of only one or two documents per cluster.

In just three days, most business units can consistently classify over 99% of documents.

3DI: From Chaos to Order

The accompanying graphic illustrates how 3DI's Graphical Analysis transforms apparent chaos into an organized structure. Each cluster can be reviewed, decisions made about disposition, and document-type labels assigned to clusters earmarked for retention.

One key benefit of the classification process is the ability to identify and dispose of non-record clusters, freeing up significant storage resources and streamlining the remaining content.

Identifying and Extracting Document Attributes with 3DI

Once documents are grouped using Graphical Analysis, the Attribute Analysis component of 3DI can extract attributes or data elements from them. This process is automatic and identifies the variable data elements in the cluster, extracting data from all members of the cluster.

After extraction, these data elements can be normalized using 3DI's Content Analysis and integrated into any CMS or business process support system. Positional cross-validation of attributes against different document types in related file groups further ensures the accuracy and reliability of data.

In essence, 3DI's combination of Graphical Analysis, Content Analysis, and Attribute Analysis offers a balanced and efficient approach to managing unstructured data. This trio delivers excellent, accurate results without needing manual intervention or data entry, transforming the way enterprises handle their data.