HBSMR Documentation

 

HBSMR WEB: Search

HBSMR WEB provides three main ways of finding information: the interactive map, free-text search, and structured search. This page describes the free-text search.

Summary

The free-text search is a google-like search of HBSMR data, content management system (CMS) pages and blog posts.

Full-text searching can operate on any or all of the main HBSMR modules (Monuments, Events, Sources, Designations, etc), and the decision around what records to make searchable is made during the build of each HBSMR WEB.  Typically there is a simple search that operates only on Monument records (plus CMS pages and blog posts), and a more advanced page for searching other modules. Results can be filtered depending on authenticated status (e.g. it is possible to have all records visible to an authenticated user while only a subset is visible to anonymous users).

Search results are displayed in a paged list and on an interactive map. The style and content of the list and map can be customised.

The free-text search box can if desired be configured to offer suggested words and phrases to the user when they start typing. This uses our "Search Predictor" component. This learns from every successful search, so every time a search finds results it will be added to the list that can be offered to the user. The display can be configured to show or hide a count, this being the count of records found when the particular search was last run, and the cache of search terms can be pre-seeded to offer users help from first launch.

Technical

The free-text searching in HBSMR WEB uses SQL Server full-text indexing operating on the XML data in the HBSMR "browser cache", plus any CMS pages and blog posts.

Because all meaningful content of each Monument record is indexed for this search, including summarised versions of linked records (Sources, Events etc), records can be found using just about any attribute, e.g. reference numbers, thesaurus terms, authors' names, as well as words occurring in site names and descriptions.

The same searching can be undertaken in API web service methods, and even through MapServer OGC services.