You must enable proper names searches before you index the data that you want to query against.
Open the HPE IDOL Server configuration file in a text editor.
Before you store content in HPE IDOL Server, terms are always stemmed and stop words are always discarded. If you want to store proper name terms (adjacent terms that begin with a capital letter) in addition to the normal content, you can set the ProperNames
parameter in the [LanguageTypes]
section to one of the following values.
Value | Meaning |
---|---|
0
|
Proper name terms are not stored. |
1
|
Adjacent capitalized terms are compounded, then stemmed and indexed as a unit. For example, Sam James is indexed as SAMJAM. |
2
|
Adjacent terms are compounded (regardless of capitalization), then stemmed and indexed as a unit. For example, bottlenose dolphins is indexed as BOTTLENOSEDOLPHIN. NOTE:
This setting considerably increases the number of terms in the HPE IDOL Server index, which can slow down its performance. |
Use the following ProperNames
options only if you need to query for proper names that contain stop words (for example, The Who or The Queen).
Value | Meaning |
---|---|
3
|
Adjacent capitalized stop words are compounded, then stemmed and indexed as a unit. For example, And His is indexed as ANDHI. Adjacent capitalized terms are compounded, then stemmed and indexed as a unit. For example, Sam James is indexed as SAMJAM. Capitalized stop words adjacent to capitalized terms are treated as individual terms. For example, The Queen is treated as THE and QUEEN, according to your stop word rules. |
4
|
Capitalized stop words are compounded with adjacent capitalized terms, then stemmed and indexed as a unit. For example, The Bells is indexed as THEBEL, and Calling Will is indexed as CALLINGWIL. Adjacent capitalized stop words are compounded, then stemmed and indexed as a unit. Adjacent capitalized terms are compounded, then stemmed and indexed as a unit. |
5
|
Adjacent capitalized stop words are compounded and indexed unstemmed as a unit. For example, And His is indexed as ANDHIS. Adjacent capitalized terms are compounded and indexed unstemmed as a unit. For example, Sam James is indexed as SAMJAMES Capitalized stop words adjacent to capitalized terms are treated as individual terms. |
6
|
Capitalized stop words are compounded with adjacent capitalized terms, and indexed unstemmed as a unit. For example, The Bells is indexed as THEBELLS, and Calling Will is indexed as CALLINGWILL. Adjacent capitalized stop words are compounded and indexed unstemmed as a unit. Adjacent capitalized terms are compounded and indexed unstemmed as a unit. |
7
|
Capitalized stop words are compounded with adjacent capitalized terms, and indexed unstemmed as a unit. Adjacent capitalized stop words are compounded and indexed unstemmed as a unit. Adjacent capitalized terms are treated as individual terms. For example, Sam James is indexed as SAM and JAME. |
You must set this parameter for each of the languages that you want to enable name recognition for (if the language settings do not include the ProperNames
parameter, you must add it). For example:
[LanguageTypes] DefaultLanguageType=English LanguageDirectory=C:\HewlettPackardEnterprise\IDOLServer\common\langfiles 0=English 1=Deutsch 2=Francais [English] Encodings=UTF8:englishUTF8 ProperNames=1 [Deutsch] Encodings=UTF8:germanUTF8 ProperNames=1 [Francais] Encodings=UTF8:frenchUTF8 ProperNames=1
Save and close the configuration file. Restart HPE IDOL Server for your changes to take effect.
Index documents into HPE IDOL Server. After you finish indexing, HPE IDOL Server treats any Query
action as a proper name query.
|