ZPORTAL at a glance

ZPORTAL provides end users with a single (federated) search environment that simultaneously searches across a selective range of information resources, and retrieves the hits into a single results set—all with the ease of an everyday internet search engine. The system preserves the richness of the facilities provided by the native resources both in their query facilities and data structures.

Benefits

• Integration of widest possible collection of resources thanks to choice of compatible data harvesting tools.
• Non-searchable resources integrated into 'one stop shop' for users
• Works with authentication services to control access to the most appropriate resources
• Efficient and seamless access to full text without further authentication
• Focused searching as resources are profiled to users' specific needs
• Options for naive and advanced users, both with intuitive user interface
• Efficient display of results ensures fast response for users
• Range of personalization features including saved searches, saved lists, filtering, sorting and de-duplication

Features

Transparent Cross-Database Searching—users can transparently search across multiple data sources simultaneously, with a single consolidated result list presented.

Cross-Domain Access—the databases searched may be of different data types. For example library catalogues, museum databases, archives, internet and intranet search engines can be integrated in to a single search interface. (Z2Web connectors may be required for some targets).

Integrated Access to Physical and Digital Resources—searches can be conducted across databases representing physical resources (e.g. library catalogues) and databases representing digital resources (e.g. e-journal databases, internet search engines).

Virtual Collections—multiple physical or virtual collections can be grouped as a virtual collection and searched as a single entity. This provides scalability for searching across very large sets of databases and provides a mechanism for providing logical groupings of database based on sector, subject content, or task, etc.

Search Profiles—System administrators create profiles to group collections such as regional resources or types of resources. End users may create their own search profiles.

Integrated Access to Internal and External Data Sources—internal and external resources can be searched simultaneously. ZPORTAL can be configured to display records from internal/local resources first.

Authentication and Access Management—system administrators can control user access to resources using a wide range of authentication technologies, including LDAP, Athens and Shibboleth.

Complete Record Display—the richness of the original data sources is maintained.

Complete Query Facilities—ZPORTAL knows the query capabilities of each underlying database, and can be configured to automatically adapt the semantics of a user's query to the "best fit" query for a particular target ("query adaptation").

High Performance—by displaying results as soon as they are returned to the server, ZPORTAL takes on the responsiveness of the fastest database in the set of resources being queried. The alternative mode of operation whereby no results are displayed to the user until all database have responded to the query leads to the system performing at the rate of the slowest system in the network.

Intelligent Holdings Display—In virtual union catalogues where holdings information is inconsistently formatted, holdings details can be configured on a database by database basis. This can also include converting institution codes into user-friendly names for display purposes.

Sort—Users can initiate a local sort on a variety of fields on the cache of results for a merged result set.

Relevance—Results are marked with 1-5 stars for relevance and can be re-sorted with most relevant items first.

Merge Duplicates—Users can de-duplicate the results returned to the server.

Links to Full Text—Embedded links are supported. Works seamlessly with optional OpenURL resolver WorldCat Link Manager and with other standards-compliant link servers.

Saved List—Records can be added to the saved list across multiple searches within a session. At any point the user can jump to the saved list to process the selected records.

Saved Search—The user may save searches to re-use in future sessions.

Search History—In addition to saving searches across sessions, the user can re-execute any search from the current session.

ILL Request—Fully integrated with the VDX ILL management system.

Integrates with Zebra and Z'MBOL—can be used with the Zebra and Z'MBOL data publishing tools to integrate access to any structured meta-data.

Integrates with Z2Web—can be used with Z2Web to integrate access to WWW-based search engines, in addition to SRU and OpenSearch targets.

Highly Scalable—ZPORTAL is designed to support searching across high numbers of databases by a large user population. The technical architecture features a parallelised CORBA environment, with a load balancing system that if necessary can share load across multiple servers.

High Performance and Error tolerance—the broadcast search engine is designed to run across large numbers of databases. The system is aware that some of the response times from databases may vary radically and occasionally database will fail to respond. Records are displayed to the user interface as soon as any server returns, hence the perceived performance of the system is that of the fastest target in the network. Databases failing to respond, or going in to error conditions have no effect on the performance of the system.

Configuration through XSL/XML—Internally all record structures are converted to their XML equivalents. These XML records are then married with appropriate XSL style sheets for each schema to generate the HTML that is displayed at the user interface. The XSL style sheets for each schema can be locally configured by the system administrator.

Profiled de-duplication—records can be de-duplicated against using a profile established by the system administrator. This profile is defined by stating which combination of data elements from a schema must match in order for records to be labelled as duplicates. When records are matched the native records are still held and can be displayed at the user interface.

Related services

• WorldCat Link Manager (optional OpenURL resolver)
• VDX (optional resource sharing and document delivery management system)
• Z2Web (optional component for searching targets that do not support Z39.50)