SAS^® Ontology Management

Features

Enables collaborative ontology development and management

• An intuitive Windows user interface provides a visual ontology structure that simplifies project development.
• Define ontologies in hierarchies of classes, and subclasses, of semantic terms to help organize the different types of information that will be accessed.
• Parent-child relationships between multiple instances associated with the same attribute (aka "slot") can be defined.
• "Slots" allow the definition of metadata that forms the definition of classes (their relevance, instantiations, etc.). Standard types include:
  • Integer (numeric).
  • Floating.
  • String (character).
  • Boolean.
  • Instance of.
  • Class ID.
  • "Slot" ID
• "Slot" constraints can include:
  • Minimum and maximum values.
  • Default and candidate values.
  • Validation against regular expressions.
  • Uniqueness.
  • Multiple cardinality.
• Allowable classes in cases of instance of, class ID and "slot" ID.
• Inverse attributes ("slots") can be defined to indicate reverse relationships.
• Instance-level and class-level versioning enable you to revert to previous definitions.
• Automatic verification of data is provided with defined relationships and inverted relationship definitions among parent-child relationships.
• Provides high scalability for users, classes and terms, and multiple ontologies.

Integrates existing document repository assets

• Upload wizards enable import and export of projects and attributes (i.e., instances).
• Nontab characters are available for separating values in multiple cardinality lists to facilitate the import and export of Excel spreadsheets, CSV files, etc.
• Built-in integration with SAS Enterprise Content Categorization is included for the automatic categorization of documents.
• Accepts XML input and creates XML output so if there are taxonomies or lists that exist or are acquired, they can be easily included as a simple import.
• Readily maps to RDF (Resource Descriptive Framework) and OWL formats.
• Output to metadata repositories, including SAS Metadata Server, SharePoint, FAST, EMC Documentum, Endeca and others, is connected via APIs.
• Data can be queried and/or modified via API calls to the server for execution of ontology applications.

Identifies relationships between document repositories

• Provides an easy-to-use interface to define relationships (i.e., semantic terms and their attributes) systematically and comprehensively.
• Organizes rules of relationships between semantic terms into classes and subclasses, which have attribute definitions that define their relevance under specific conditions. Relevance conditions (otherwise known as instantiations) include definitions about the associations, characteristics and constraints specified by the user.
• Characteristics and constraints comprise the metadata relationships and form the definition and operation of the class.
• Rules associate text documents and repositories by their class definitions; not just the key words contained in the materials.

Builds subject-matter expertise into search-and-retrieval activities

• Lets you create an ontology that establishes and organizes the hierarchical relationships of semantic terms using advanced linguistic technology.
• Identifies the explicit and implicit connections (i.e., structural relationships) between document attributes. Identifies classes and subclasses of information within hierarchies.
• Lets you define attributes associated with each class and subclass within the hierarchy.
• Automatically applies instantiation definitions to appropriate content when information is sought through search or retrieval, regardless of document location or language.

Consistently applies subject-matter expertise in real time

• A predetermined list of semantic terms helps organizations begin their ontology development.
• Easy-to-use GUI enables users to add semantic terms to customize relationships among terms.
• Relevance and instances of occurrence can be specified to consistently identify relationships among areas of interest. The resulting hierarchies can be used as input taxonomies to automatically categorize content using SAS Enterprise Content Categorization.
• Content can be associated per the defined ontology at run time, creating metadata when the ontology is connected to a database using the product APIs.
• Advanced browsing of data is supported so that key document attributes can be clearly understood.
• A query window supports both simple and complex (linked simple queries) to find desired classes, drill into their details and replace associated information as needs change.
• Hierarchy and inverted hierarchy views facilitate parent-child relationship understanding and can be used as taxonomies for downstream applications.
• The ontology management server provides the ability to show users the results by applying the ontology to document collections in real time.

Centralizes administration for collaborative ontology development and maintenance

• Ontologies are retained as projects that can be saved, shared and collaborated on for ongoing maintenance and updates, with users' rights controlled by administrative authorizations.
• Default permissions can be set at each of the three project levels.
• Administrators can specify individuals as members of groups with collectively defined permissions.
• Users can be limited to read-only access, or they can have the ability to add, delete or modify classes, attributes (aka "slots") or instances.
• Supports class- and instance-level version control.
• Provides find-and-replace capabilities to retain consistency with broad changes.
• Provides report logging that records what instance-level changes were made by whom as part of the standard audit functions.
• Semantic terms are systematically managed and the conditions and connections associated with isolated types of key information are defined so those defined as most relevant are delivered to the right person.