Extending taxonomy efficacy

Convera has unveiled its new Taxonomy Workbench, which allows multi-national companies to meet the new organizational mandate that information must be accessible to, and shared with, all users—no matter where that knowledge resides or in what language.

The Taxonomy Workbench tackles the business challenge that global intranets must provide intelligent, reliable, consistent information—in multiple, translated languages. It facilitates electronic collaboration through those intranets while eliminating the additional resource costs commonly incurred with what Convera calls frequently error-prone automatic taxonomy generation software, which requires substantial downstream taxonomy adjustment and system training.

As Convera explains it, the Workbench helps deliver more accurate search results within corporate intranets, providing geographically scattered employees quick and easy access to diverse information--or on an external Web site, for researchers, customer service and product information.

A complete set of tools for organizing large amounts of disparate data, says Convera, the Workbench reduces the time to build effective taxonomies from months to days and is designed for easy use by information managers, librarians and subject matter experts.

The company explains the new product makes taxonomy creation easy for a broad range of users by:

Developing completely new taxonomies. Customized to a corporation’s particular needs, the Workbench helps create an easily searched organization structure for each organization’s own specific online public information.

Importing existing taxonomies and thesauri. To leverage data organization models that are already in place, users can integrate and modify existing taxonomies or thesauri. Automated import, generation, combination and pruning wizards aid in accelerated taxonomy and classification development.

Testing, benchmarking and tuning taxonomies and classifications. A full suite of taxonomy quality measurement tools helps ensure that employees looking for information on a corporate intranet or the general public using an organization’s Web site will not need to search unnecessary layers of data for details they require. Convera’s static and dynamic benchmarking tools are said to significantly reduce the time to create and maintain high-function taxonomies, offering measurable ROI. Utilizing the graphical interface, users can highlight certain data, identify taxonomies and simply watch their monitor as the newly created taxonomies are populated by dispersed data. All categorization processes, including normalization, latching, idiom detection, disambiguation and ranking are performed through benchmarking.

Translating human subject area expertise into fast Web site browsing and searching. User friendly and well-suited for large organizations, the Taxonomy Workbench allows organizations to capture the expertise of domain subject experts online by building in-depth methods of search on a particular data set. Multiple users on different development platforms can work simultaneously, yet securely, creating taxonomies and classifications.

In other Convera news, the company has announced thorough and user-focused taxonomies in genetics, finance and business, general enterprise, technology and U.S. government. Convera reports these taxonomies contain industry standard thesaurus content enhanced by Convera linguists and taxonomists to optimize the quality of information discovery in government and commercial applications.

The company says each of the new taxonomies speeds the process of organizing an enterprise’s knowledge assets by incorporating specialized vocabulary and ensuring the most accurate level of categorization available, explaining that large and diverse collections of unstructured data can be efficiently browsed by navigating categories that users find intuitive and useful to their personal information retrieval objectives.

Convera's linguistic taxonomy approach delivers the necessary building blocks for organizations to develop a unified knowledge framework from controlled vocabularies, authority lists and classifications.