A uniform resource identifier (URI) is a unique string of characters which identifies and directs users to a particular area of content on the Internet. URIs of particular relevance are indexed and connected to market areas for optimum interrelation of the widest possible range of useful data.

Each of these fields holds the address of one named site, so a link belongs in the field for the site it points at and nowhere else. A page about the building on a news site or a developer's own page has no field here and is recorded as a source instead.

The link has to reach the page for this record and not a search result, a category listing or the site's front page. A link to a list that mentions the building is not a link to the building.

Where the site also issues an identifier, the two are kept separately. The identifier survives a site reorganizing its addresses, and the link does not, so a record carrying both stays usable when one of them breaks.

A link that has stopped working is still evidence of what the record was matched against, so it is worth correcting rather than deleting.