Manage Smart Search Knowledgebases

To add a new Smart Search knowledgebase, follow these steps:

1. From the System Navigator, click System Administration > Ongoing Maintenance > Smart Analytics > Smart Search.

   The Smart Search Configuration page opens.

2. From the Add Knowledgebase section, type a unique name for the new Smart Search knowledgebase in the Knowledgebase name field (required).
3. Select the knowledgebase type from the Type field drop-down list.
4. Click Add. The new IDOL knowledgebase record is added and a new knowledgebase maintenance page is displayed.

5. Configure the required knowledgebase information. The field availability varies for different knowledgebase types.

   Field	Description
   Display Name	Types a display name for the IDOL knowledgebase.
   Refresh Interval	Displays the number of interval unit used to update the selected knowledgebase index. One interval unit is 5 minutes. Default number of interval is 1. You may increase the interval, which extends time delay between updates, by increasing this number. Setting this number to 0 (zero) disables updates to the index. To re-start indexing, reset the interval to a value greater than zero.
   Connector (not for sclib)	Selects the specified connector from the configured connectors according to the knowledgebase type: weblib: HTTP connector splib: SharePoint connector fsylib: File System connector The connector selected here should be the be consistent with the configurations in the .cfg files of related remote connectors. For more information about configuring Smart Analytics connectors, see Configure connectors.
   Task (not for sclib)	Selects the task that connects the Smart Analytics server and other data sources. The task selected here should be consistent with the configurations in the .cfg files of related remote connectors. For more information about Smart Analytics connectors, see Configure connectors.
   Table Name (only for sclib)	The table that will be indexed. A valid Service Manager table is required.
   Skip These Extensions (only for sclib)	A semicolon-separated list of file extensions that should not be indexed or extracted. Certain file types either cannot be indexed, or provide no relevance. By providing these extensions, you can increase index performance. Note Sample gif;jpg without any spaces.
   Document ID Field (only for sclib)	Every table in Service Manager has a unique ID field and this field identifies the field name of the ID field. The indexer uses this field to uniquely identify each document in the index. This is a required field for indexing a knowledgebase.
   Index Attachments (only for sclib)	If the table being indexed has attachments, select this check box to have the attachments indexed.
   Max threads for full reindex (only for sclib)	Specifies the number of threads to process the records in the library for Smart Serach indexing. The default value is 1, which means the system creates one back-end process (worker) to send the data in the library from Service Manager to Smart Analytics. Note For performance considerations, we recommend that the total number of threads for all the libraries is less than 20.
   Table Query (only for sclib)	Specifies a Service Manager style query to limit what records in the table are indexed. For example, a query to return only documents that are neither draft nor retired in the kmdocument table: status ~= "draft" and status ~="retired". A blank query indicates that all records will be indexed.
   Status
   Full Reindex:	When clicked, the search engine performs a full re-index of a knowledgebase. If the index does not exist in IDOL, an index will be created. If the index does exist, Smart Analytics replaces the existing database with the new index.
   Refresh Statistics	When clicked, the search engine refreshes the statistics for this library.
   Field Definitions (only for sclib)
   Field Name	Specifies the field name in the Service Manager table to be included in the index.
   Alias	Specifies the name that the field is to be indexed as.
   Type	Indicates whether the field is a plain text string, a rich text string, or a date type. The indexer ignores HTML markup in rich text strings and indexes plain text strings completely. Rich text: When the field type is Rich Text, it means the field may contain HTML code. HTML tags will be removed during index.. Reference: Select Reference only when the current field is a CI logical name. Thus the Display Name of a CI can be indexed.
   Display in Results	Defines what fields are available in the search results. Fields marked as "true" in this column are available to be included on a search results. The information for these field is displayed at the bottom of the search result.
   Index weight	Defines the sequence of the search results which are displayed on the search hitlist. You can assign different index weight to adjust the display sequence. If the sort value of a field is assigned as level 4, the search results of this field is displayed at the top of the list. No index means that this field is not used for search. Note HPE recommends you to tune this field according to your own business use cases.
   Match	By setting the field in this column to "true," the system indexes the field's content as an advanced search filter for Smart Search in this knowledgebase. The search engine searches this individual field with the Advanced Search filter. For example, if you set the Category field as true, this field is used as an advanced search filter. If you leave this value as empty, it applies the false value.
   Sort	Defines the sequence of the search results which are displayed on the search hitlist. If you set multiple fields as hitlist, you can decide the display sequence by assigning different sort values for these fields. If the sort value of a field is assigned as 1, the search results of this field is displayed at the top of the list.
   Data cleasing	Used to remove text which matches the data cleansing rules from the table before indexing.
   Script
   Knowledgebase access script	This script specifies the script the system uses to determine if a particular user has rights to access the knowledgebase. See the default script for detailed information.
   Category index script	This script processes the document category so that the indexer can translate the document’s category into a string that the search engine can use later to find the document based on the user’s category access.
   Search security script	This script is used to build and return a string of library-specific query values that were entered by the user under the tabs in the Advanced Search screen. Tailor this script when a knowledgebase has a tab in the Advanced Search screen and you wish to modify the fields available for Advanced Search.
   Default Locale	Specifies the default language used by the search engine when searching and indexing. By default, the language code is English.

6. Click Save to save your modification.