P@noptic 5.0.0 release notes

Released: 5 Nov 2003

Prior to v6.0 Funnelback was known as P@noptic Search

Panoptic V5.0 addresses a significantly larger enterprise search market. It is now available in Solaris and Windows versions as well as Linux and supports indexing of databases (Oracle, SQL Server, MySQL). Panoptic 5 also provides even easier customisation, even easier integration with other systems, even higher quality results and even more search features.

Broader applicability

  • Windows 2000 and Solaris versions (in addition to Linux).

  • Database indexing - Oracle, SQL Server, PostGres, MySQL etc.

  • Crawling of HTTPS sites.

  • Crawler support for cookies and session-id stripping allows even better crawling of dynamically generated websites.

  • Highly customisable structured search of XML formats.

  • Highly customisable metadata configuration. Netscape, Dublin Core, AGLS and user-defined metadata schemes.

  • Support for media formats such as JPEG and CMML (See the Continuous Media Web site).

  • Support for alternate single-byte character sets such as ISO-8859-2, Windows-1251 and koi8-r allow search of languages such as Russian, Polish, Norwegian, and Greek. Character set mappings are customer-definable if an existing mapping isn’t available.

  • Accent conflation - an unaccented query word will match all the accented variants. For example, chateau will match ch√Ęteau. (This feature can be disabled if not required but allows searchers with simple keyboards to conveniently search for foreign language text.)

  • Improved documentation on integrating Panoptic with other systems.

  • Even higher quality ranking of search results:

    • Better elimination of near-duplicates.

    • New high-performance ranking algorithms based on new CSIRO research.

    • Configurable down-weighting of results from the same "site"

    • Links are redirected to ensure proper targeting and counting of link information

  • W3C double-A accessibility compliance for search pages.

  • A full range of query refinement options, including search within results, stemming, spelling suggestions, sounds-like and customisable thesauri.

  • Improved summaries including punctuation.

  • Choice of summary formats. snippet, metadata, query-biased and combination modes selectable on a per-query basis. Summary length also controllable for individual queries.

  • Much greater control of "relevance" scoring. (If you’re sure you can do better than the defaults!). For example, you can select a group of metadata classes which you wish to be upweighted.