The way that the name is combined before it is passed to NAMESET or used in MATCHing determines how the following parameter should be set:

NAME-FORMAT=(L or R)

The

NAME-FORMAT

parameter tells SSA-NAME3 where to commonly find the major word in the name field, at the left end or the right end. The major word in a Person name is usually the family name. The major word in a Company Name tends to be the left-most significant word. The major word in an Address is usually the Street name. SSA-NAME3 uses its knowledge of the major word position for positive searches, if only a single key was requested for the name, and to allow weighting of the major word during the matching process. For example,

NAME-FORMAT=R

is a common setting for Western person names.

NAME-FORMAT=L

is a common setting for Chinese names. If your data contains a mixture of name types, and the major word is not in a stable position, then you should use multiple keys, not a single key, and then not bother to weight the major word in matching. SSA-NAME3 uses a tiered approach to discovering the major word in the name, and the

NAME-FORMAT

parameter setting is at the bottom of the tier. Firstly, if SSA-NAME3 finds a comma (,) in the name, the word preceding the comma is taken to be the major word. This comma processing can be turned off by setting

CLEANING-OPTIONS #2

to

Y

. Secondly, Major ’Markers’ can be defined in the Edit-list. These can be in the form of special characters (e.g. / or @) or in the form of words (example, ST, RD, RUE or JALAN for the determination of a Street Name major word). The word adjacent to a Major Marker is taken as the Major Word. For more information on Major Markers refer to the

Edit-list

chapter of this guide, and the

Cleaning

chapter of the

APPLICATION REFERENCE FOR SSA-NAME3 SERVICE GROUPS

guide. If no major word has been identified, then the setting of

NAME-FORMAT

is used. For example, when processing addresses, if

NAME-FORMAT=L

and SSA-NAME3 finds no street name major marker in the address, it will go to the left end of the address and search for the first non-delete, non-skip word. Delete words (such as THE, C/O) and Skip Words (such as APARTMENT) can be defined in the Street Edit-list and are customizable.

Numbers can also be defined as skip words.

NAME-FORMAT=L

would be an appropriate setting if the whole address was passed to NAMESET for key building, so as not to look at the locality end of the address.

NAME-FORMAT=R

would be an appropriate setting if only the part of the address preceding the locality/state was passed to NAMESET for key-building, in case passed string contained no Street Type.