eClassify: Enterprise Knowledge Assets management system with automated file classification support

Start Date: 01/04/2004
End Date: 31/03/2006
Funding: New forms of organisation (GSRT – national programme)
Project Leader: Piperidis Stelios

Knowledge management is the set of systematic and disciplined actions that an organization can take to obtain the greatest value from the knowledge available to it. Knowledge in the context of the eClassify project includes both the experience and understanding of the people in the organization and the information artifacts, such as documents and reports, available within the organization and in the world outside. The user of the project is Piraeus Bank Group, one of the most dynamic financial organisations in Greece today. The technology that has been developed in the framework of the project falls into the domain of text classification and indexing, assisted by the taxonomy that underlies the Bank’s intranet structure. The main objectives of the project were:

  • Better organization and lower maintenance costs of the user information.
  • The development of an efficient text categorization algorithm functioning on the taxonomy of the user, that categorizes new documents with high accuracy and produces an index for retrieval purposes.
  • The integration of the categorization engine with the user content management system on top of which are built the information portals.
  • Analysis of the intangible assets of the user, its knowledge capital and its structured and unstructured content.

The market for classification software is dynamic and evolving very fast. Many companies are supplying this software. Many approaches and technologies have been proposed in the bibliography. Methods that are used include neural networks, bayesian analysis, linguistics and semantics, SVM, pattern matching and other statistical algorithms. Also, there exist many approaches to building and populating the taxonomy. Most of the vendors view this technology as fundamental in the information infrastructure. The design philosophy positions taxonomy and classification as a core product in a broader suite of products that work on the unstructured data of an organization. Major vendors are: Autonomy, Convera, Entopia, Mohomine.

Addressing enterprise knowledge assets by categorization and search is the first step towards cashing in on its information assets. Knowledge workers can find the information they seek and quickly leverage the investment made in enterprise content, without spending time recreating information that already exists. For enterprises like Piraeus Bank, user efficiency and effectiveness has been increased substantially. Sales and revenue increase, customer and support costs decrease. Adding new content to the intranet information portal is greatly simplified.