Setting Up Search Collections

This chapter provides an overview of PeopleSoft CRM searching and discusses how to:

• Define thesaurus and noise words.

• Define record-based indexes.

• Define HTTP-based indexes.

• Define file-based indexes.

• Configure search options.

• Build and test a search collection.

• Define jobsets for the build collection process.

• Use the FTP option to index files.

• Perform search on the universal search tool.

Understanding PeopleSoft CRM Searching

This section discusses:

• Search collections and their uses.

• Solution and error searching.

• Enhanced search capabilities.

• Search index templates.

• The Build Collection process.

• System configuration.

• The universal search tool.

Search Collections and Their Uses

Several PeopleSoft CRM transactions incorporate a search that is powered by Verity, a third-party search and retrieval tool. The system searches against a Verity collection, a separate set of files that contain data that is copied from the PeopleSoft CRM database. The subset of data that is included in a collection is based on search index templates that you define.

PeopleSoft CRM includes these collections:

• The product catalog collection (CRM_RB_PRODCAT).

  This collection is used to search for products that you can order. All users can search the product catalog from the Lead, Opportunity, Order, and Quote components and directly from the menu. Users cannot add products to an order unless they license PeopleSoft Order Capture.

  Note. If Order Capture is not licensed, disable the Add to Order buttons using the Catalog template.

  See Creating Catalog Display Templates.

• The unstructured email collection (CRM_RB_ERMS).

  This collection is used by the email response management system (ERMS) in PeopleSoft Multichannel Communications. Text from inbound emails is stored in the collection and then scored against keywords that you define. ERMS uses the scores to route the email to the appropriate worklists so that the most qualified agents can review and respond to the email.

• The solution and error message collection (CRM_RB_SRCHDB).

  This collection is used by the self-service PeopleSoft Support application to return matching solutions for cases.

• The defect and fix collection (CRM_RB_DEFECTFIX).

  This collection is used by the PeopleSoft Quality applications to return matching defects and fixes.

• The self-service case collection (CRM_RC_CASE_SS).

  This collection is used by the self-service PeopleSoft Support application to return matching solutions for cases.

• The solution collection for the PeopleSoft Integrated FieldService application (CRM_RF_SRCHDB).

  This collection is used by the PeopleSoft FieldService application to return matching solutions for service orders.

• The email workspace collection (ERMS_SRCHDB).

  This collection is used by the email workspace to return matching solutions and templates (for sending response) for incoming emails.

• The chat collection (SEARCH_SOLUTIONS).

  This collection is used by the chat application to return matching solutions for questions that are entered in chat sessions, and by the PeopleSoft Quality application to return matching solutions for defects.

You must update the product catalog collection and the solution and error message collection regularly to ensure access to current data. The system automatically updates the unstructured email collection every time it processes newly arrived emails.

See Also

Solution and Error Searching

Understanding Product, Pricing, Proration, and Catalogs

Process Instantiation

Solution and Error Searching

You can use the solution and error message search collection in various ways.

Solution Advisor for Cases

In the three PeopleSoft CRM call center applications (PeopleSoft Support, HelpDesk, and HelpDesk for Human Resources), Solution Advisor is used to search for objects that can help resolve a case. Search groups determine which objects are searchable in agent-facing and self-service versions of Solution Advisor:

• The agent-facing version searches solutions; it also searches cases and troubleshooting scripts, both of which can lead the searcher to a solution.

• The self-service version searches only solutions.

Both versions enable users to associate selected solutions with the case from which the Solution Advisor was invoked. The self-service version additionally enables self-service users to search for solutions outside the context of a case.

Solution Advisor for Service Orders

In PeopleSoft Integrated FieldService, Solution Advisor is used to search for solutions that are relevant to a service order. PeopleSoft Integrated FieldService does not track solution usage for service orders; search results are informational only.

Similar to Solution Advisor for cases, the type of object that gets searched (solutions only) is controlled through search group configuration. As delivered, Solution Advisor for service orders searches only solutions.

Solution Management Searches

Two solution management operations incorporate searches:

• Merging similar solutions.

• Automatically relating similar solutions.

Both processes use searching to find solutions that are similar to a base solution that you indicate.

Error Validation Searches

Error validation searches enable PeopleSoft Support and PeopleSoft HelpDesk users to search for known error messages that are similar to an error message that callers report. This helps users match reported information to a known error even when callers provide incomplete or misspelled information about the message.

This functionality is available only if error tracking is enabled for the call center business unit.

See Also

Managing Solutions

Solution Search Templates

Processing Cases

Working with My Service Orders

Defining Call Center Business Units and Display Template Options

Enhanced Search Capabilities

Enhanced search capabilities enable you to provide the user with a robust, knowledge management solution.

Enhanced search capabilities include:

• HTML indexing

• File indexing

• Thesaurus

• Word variations

HTML Indexing

You can configure Verity search to include external content using the web through the use of the spider/crawl HTTP indexing capability. This feature allows data to be collected from external websites. The spidering capability indexes the structure of the site and builds a reference index that becomes searchable through Verity. These HTML files become searchable through Solution Advisor, without physically pulling the content into the Verity collection. The system returns results in the form of URL links with content summaries for those pages in the search results. The system displays the stored summaries in cases in which the referenced URL is no longer accessible.

When a customer service representative (CSR) is presented with a list of links to external content, the CSR can navigate to the external content and view the content on the remote system. If the CSR believes that the content is relevant and useful to the case, the CSR can choose to use this information as they would an attempted solution. External content can be converted into a new solution containing a complete copy of the text of the page as it existed at that moment in time.

If the CSR deemed the information not relevant or useful to the problem, the external content is not converted into a solution and no associated content is brought into the solution database. External web content is converted into a solution the first time that an agent finds the information valuable to a case. For all future searches in which the content of this page is deemed relevant by Solution Advisor, the system returns the solution and not the corresponding external link.

File Indexing

You can configure Verity to include collections that reside on file servers or directories. The system uses Verity's file system capability for the spider/crawl index to gather data from select files or directories (for example, a network drive share) and return links and summaries of those files in the search results. This enables agents to search for documents and other files that are related to the search criteria. The agent can create a relationship between the selected file and the transaction (case) to retain usage characteristics and track the document's contribution to the successful resolution of the customer issue. When an external file is deemed relevant and attempted for the first time, it can be converted into a new solution so that it can be associated with the case.

Thesaurus

You can configure the Solution Advisor search to consider alternate words for one or many of the words in the search field. The accuracy and relevancy of the search results are improved when alternate words are considered in the search. For example, a search criteria that included the word PC would also consider terms such as desktop, laptop, personal computer, and computer in returning a set of solutions or other results that are related to the entered search criteria. You can create custom thesaurus for the keyword search, use the thesaurus that Verity delivers, or both.

Noise words

You can configure the Solution Advisor search to ignore noise words in order to reduce the number of matches on trivial words from the search text. Examples of possible noise words might be and or for.

Word Variations

Solution Advisor can be configured to allow a number of word variations, including the ability to search against alternate spellings of common words, search for words with similar sounds, search for derivative variations, and so on. Users can optionally disable spell check if performance suffers by the introduction of this feature.

Search Index Templates

Search index templates control the records and fields that are included in a collection.

Search Index Template Configuration

You create a separate search index template for each table or view that is included in a search collection. For example, users of Solution Advisor for cases need search index templates for cases, solutions, and troubleshooting guides.

PeopleSoft delivers search index templates for the search collections that are described in the preceding sections. You can modify these templates to change the fields that are searched, although it is not normally necessary to do so.

Search templates can include an optional SQL where clause to filter out data that you don't want added to the search collection. This improves the performance of the process that builds the collection. They also include field-level options that determine which fields are searchable and which fields are returned to the originating application.

You can improve performance for both the collection build process and for searching by:

• Including fewer fields in the search collection.

• Avoiding long character fields such as comments.

You can also improve the efficiency of the build process by basing search index templates on views that contain only the fields to include in the collection. In the background, the PeopleCode is carrying out a %SelectAll command on the record, so if fewer fields are in the record, the build is faster.

Delivered Search Index Templates

This table lists the records for which PeopleSoft delivers search index templates for the product catalog collection (CRM_RB_PRODCAT):

This table lists the records for which PeopleSoft delivers search index templates for the solution and error message collection (CRM_RB_SRCHDB):

This table lists the record for which PeopleSoft delivers search index templates for the unstructured email collection (CRM_RB_ERMS):

This table lists the records for which PeopleSoft delivers search index templates for the defect and fix collection (CRM_RB_DEFECTFIX):

This table lists the record for which PeopleSoft delivers search index templates for the self-service case collection (CRM_RC_CASE_SS ):

This table lists the records for which PeopleSoft delivers search index templates for the solution collection for PeopleSoft FieldService (CRM_RF_SRCHDB):

This table lists the records for which PeopleSoft delivers search index templates for the email workspace collection (ERMS_SRCHDB):

This table lists the record for which PeopleSoft delivers search index templates for the solution collection for chat (SEARCH_SOLUTIONS):

The Build Collection Process

After you set up index templates for collections, you use the Build Collection process (RB_SRCH_BLD) to build the collection.

Building the Product Catalog Collection

When building the Product Catalog collection, run the Load Catalog Cache process (RO_CAT_PUSH1) before the Build Collection process.

The Product Catalog Collection job (BLD1CAT) incorporates both required processes.

Building the Unstructured Email Collection

You do not explicitly build the unstructured email collection. Instead, the Unstructured Content Analysis job incorporates both the Build Collection process and the process that analyzes and routes newly arrived inbound email.

Building Collections for External Contents

To search for information that exists outside the CRM system, such as external web pages and file systems of the enterprise, you need to build search collections for them as well. The procedure is the same as setting up a record-based collection. In the run control definition of the Build External Content process (RB_BLD_EXTCT), specify the type of external content (file or HTTP) and the index definition that is used to build the search collection.

Building Collections on Multiple Application Servers

The search collection resides on an application server. If you have multiple application servers, build the collection on each one. Having each application server access the collection files locally ensures optimum performance and simplifies the application server configuration.

You can manage multiple application servers in two ways:

• Build the collections on one application server at a time.

  When you do this, the build process runs on the PeopleSoft Process Scheduler server (and therefore on the application server) that you select on the Process Request page.

• Build collections on all application servers at once.

  To do this, set up a PeopleTools JobSet that initiates the processes on multiple PeopleSoft Process Scheduler servers.

  See Defining JobSets for the Build Collection Process.

Create and Update Modes

The Build Collection process has two modes: create a new collection and update the existing collection. The Update mode is faster, but it has certain limitations. Be sure to use the Create mode the first time that you build the collection and after you create a new search index template.

When collections are built in Update mode, the collection is totally locked by Verity This is called an exclusive lock. When an agent tries to access a collection that is in an exclusive locked mode, it may take longer to get results back. Also, doing an update forces Verity to maintain the old document as well as the updated document, thereby increasing the document file size.

To always keep a copy of the collection online while it's being built in Create or Update mode, the system uses a temporary name for the collection that is being built, leaving the current collection available to users during the build process. When the build is complete, it exchanges the active collection name with the temporary collection name so that the newly built collection is the one that is used for searches. The older version of the collection remains on the system, but only the newly built one is accessed during searches.

If the Build Collection process is run in update mode and no new data is added to the collection, the process finishes with a success status (log information is available in the message log) instead of returning an error message. It ensures that the update process doesn't get interrupted by the error, especially for customers who set the Build Search Collection process (RB_SRCH_BLD) to recur regularly.

Steps in the Process

The Build Collection process performs this sequence of steps:

1. Accesses the setup files, the style files, and the index templates.

   PeopleSoft CRM leverages the style files that are delivered by PeopleTools so that no additional CRM-specific style files need to be installed separately. To make the style files work for CRM, the Build Search Collection modifies the style.uni file when it runs on a particular PeopleSoft Process Scheduler server for the first time.

2. Creates the BIF (binary interchange file) and DAT (data) files for the data rows that meet the Where clause criteria.

   An incremental update gathers changes to the index templates and new information that was added to the database since the last collection update. These new rows are used to create new BIF and DAT files that are added to the collection. This creates a new partition within the search collection and makes the data available in the search results.

3. Builds a BAT (batch) file or shell script that copies the Style.ufl file as well as other style files into the collection, based on language code.

4. Runs the MKVDK executable program that builds the collection.

   The file is located in the PS_CFG_HOME\appserv\prcs\database\files directory in Microsoft Windows and in the PeopleSoft Process Scheduler home directory in UNIX.

5. Updates the Date/Time field to indicate when the collection was created or updated.

6. Resets the usage count and the last usage Date/Time field if you selected the Reset Usage Count check box for the build.

7. After the process is finished (succeeded or failed), the log information of the MKVDK program is captured in a file called VerityOutput.Dat; the log is posted in the report repository. The log information is also displayed in the message log and the redirected terminal output on the View Log/Trace page. The content of the new collection is available in the <search collection name>.dat file on the same page. The same log process is also available to the collection building process for external context.

Note. Most search transactions do not track usage counts. Solution Advisor does count usage for solutions, but it is based on associating the solution with a case, not on the usage count fields that you set up in Verity. If you accidentally reset usage counts, the PeopleSoft call center applications provide an update usage count process that can recount solution usage.

When the Build Collection process is complete, you can find the collection on the server's Verity search collection path in the directory with the same name as the collection in the appropriate language-specific directory.

Performance and Optimization

To reduce the number of partitions in the database, configure search settings so that the collection is optimized on a regular basis. An optimization maximizes the size of the partitions that are in the collection. In most cases, optimize the collection after five incremental updates. Because the optimization frequency is set on the Search Settings page, you do not have to do anything when you run the build process.

Factors that affect performance of the Build Collection process include:

• The amount of data that is processed.

  Reducing the amount of data that is processed reduces the time that is required to build the collection. You can limit the data that is included in the collection by including a Where clause in the search index template definition. Also, running incremental builds instead of full builds can reduce the amount of data that is processed during a build.

• The speed of the computer that runs the Build Collection process.

  Because the Build Collection process is CPU intensive, run it on a computer with a fast processor.

• The amount of memory that is on the computer that runs the Build Collection process.

  The computer should have at least one billion bytes (technically 1,073,741,824 bytes) of memory.

Errors During the Build Collection Process

The build process fails under these conditions:

• PS_HOME is not set or not on the PeopleSoft Process Scheduler path.

• The search collection path that is specified on the Build Collection page does not exist or is not accessible from the PeopleSoft Process Scheduler server.

  The path that you enter on the run control page is relative to the application server, not to the computer from which the request is made (PeopleSoft Process Scheduler is assumed to be running on the application server computer). The search collection path must be outside of PS_HOME.

• The Verity executables were not installed.

  Verity is installed as part of PeopleTools.

• The PeopleSoft Process Scheduler does not have administrator ability to create or delete files.

System Configuration

This section discusses system configuration considerations.

PeopleSoft Process Scheduler Configuration

When you initiate the Build Collection process, you select the PeopleSoft Process Scheduler server that will run the process and provide a path to the collection files on the application server. Because the path is relative to the PeopleSoft Process Scheduler server, the setup is greatly simplified if you set up a PeopleSoft Process Scheduler server on each application server.

This configuration means that whenever you build a collection, you provide the same (local) path regardless of which PeopleSoft Process Scheduler server runs the process. It also ensures that the PeopleSoft Process Scheduler server has access to the disk where the collection files reside.

Note. We strongly recommends that the PeopleSoft Process Scheduler server and application server reside on the same server. If you have multiple application servers, we recommend that you set up a PeopleSoft Process Scheduler server on each one.

A PeopleSoft Process Scheduler that runs the Build Collection process must use a sign-in ID that has administrative abilities to create and remove the files and directories that are created as part of the Build Collection process.

To enable PeopleSoft Application Engine programs that call the Search API to locate collections, add the following section to the PeopleSoft Process Scheduler configuration file, making sure to modify it with the appropriate collection name and path.

[Search Indexes]
CRM_RB_SRCHDB=c:\pt880\data\search\cr880dvl
;======================================================================
; Search index settings
;======================================================================
; Search indexes can be given alternate locations if there is an entry here.
; Entries look like: IndexName=fs location (ie EMPLOYEE=c:\temp)

Please refer to the Supplemental Installation Instructions for PeopleSoft 9.1 Applications (available in My Oracle Support) for step-by-step instructions on how to set up Verity searching in PeopleSoft CRM.

Application Server Configuration

Each search collection must have an entry in the Search Indexes section of the application server configuration file (PS_CFG_HOME\APPSERV\DomainName\psappsrv.cfg). The following example shows entries for the three PeopleSoft CRM collections.

[Search Indexes]
CRM_RB_PRODCAT=C:\VERITY\DATA\SEARCH\DatabaseName
CRM_RB_SRCHDB=C:\VERITY\DATA\SEARCH\DatabaseName
CRM_RB_ERMS=c:\VERITY\data\search\DatabaseName

For Windows and UNIX operating systems, make sure that the PS_HOME environments variable points to the current tools home directory; for example, PS_HOME= c:\pt880 on Windows, or /ps/psasgrp/casb on UNIX. The batch or shell file that builds or updates the collection uses this variable. It should not contain an ending slash or backslash.

For UNIX, make sure that the Verity executable programs are part of the path-defaulting set of directories. Make sure that the MKDVK executable is part of the path. You must also make sure that chmod and sh are located in the home directory for the PeopleSoft Process Scheduler server or set a soft link pointing to those commands. For example:

ln -s /bin/chmod $PS_HOME/chmod
ln -s /bin/sh $PS_HOME/sh

See Also

Supplemental Installation Instructions for PeopleSoft 9.1 Applications

The Universal Search Tool

When you use the universal search tool, you first select the type of object that you're looking for; values are based on the objects for which you have defined search index templates.

Once you've selected a record, there are two ways to define search criteria. You can use either method by itself, or you can use both of them together.

You can enter record-level search criteria. The system searches for matching text in all of the fields included in the record's search index. The system ranks the search results according to how close the match is. If you enter record-level search criteria, you can use a natural-language search or a keyword search.

Alternatively, you can enter field-level search criteria. After you select a search record, the system displays a list of searchable fields. For each field, you can enter a search operator and search text. You can also decide whether to limit the results to records that meet all field-level criteria or whether to accept records that meet any individual field-level search specification.

This table describes each of the search operators available for field-level searching:

Note. Searching is not case-sensitive.

Defining Thesaurus and Noise Words

To define thesaurus and noise words, use the Manage Thesaurus (RB_MNG_THESAURUS) and Manage Noisewords (RB_VRTY_STOPWORDS) components.

This section discusses how to:

• Load the Verity thesaurus and noise words.

• Define a custom thesaurus.

• Work with a delivered thesaurus.

• Manage the thesaurus.

• Define custom noise words.

• Work with delivered noise words.

• Manage noise words.

Pages Used to Define Thesaurus and Noise Words

Loading the Verity Thesaurus and Noise Words

Access the Load Verity Noise Words and Thesaurus page (Set Up CRM, Utilities, Search, Load Verity System Files, Load Verity Noise Words and Thesaurus).

Specify the type of data that you want to load from files that are provided by Verity into the CRM database: thesaurus, noise words, or both. Based on the data and language selection, the AE process (RB_LOAD_VRTY) reads the respective Verity files and loads data into the CRM database.

This is not a mandatory step. Run this AE process only if you want to take advantage of the thesaurus and noise word lists that Verity provides.

Defining a Custom Thesaurus

Access the Custom Thesaurus page (Set Up CRM, Utilities, Search, Manage Thesaurus, Custom Thesaurus).

Synonym For	The word for which the synonym will be defined.
Synonym List	Enter a list of possible synonyms for the word. Separate the synonyms with a comma.

Working with a Delivered Thesaurus

Access the Delivered Thesaurus page (Set Up CRM, Utilities, Search, Manage Thesaurus, Delivered Thesaurus).

The system displays all thesauruses that are supplied by Verity and loaded on the system. Deselect or delete the ones that you don't want to use in the verity search. To delete the entire thesaurus, click Select All and then click Delete All.

Note. To improve performance when a large delivered thesaurus is loaded, PeopleSoft limits, by default, the display of delivered thesaurus synonyms to the first 100 entries. To view all of the entries, click the View All Delivered button. Clicking this button, however, can slow down system performance.

Managing the Thesaurus

Access the Manage Thesaurus page (Set Up CRM, Utilities, Search, Manage Thesaurus, Manage Thesaurus).

Select the types of thesaurus that you want to view. The system displays a list of entries for each type of thesaurus that you select.

Defining Custom Noise Words

Access the Custom Noisewords page (Set Up CRM, Utilities, Search, Manage Noisewords, Custom Noisewords).

Enter the words that you want the system to identify as noise words. Separate words with a comma. Select the language code. Click Add Noise Word(s). The grid displays all the noise words that you added.

Note. All custom noise words should be entered using lower case letters.

Working with Delivered Noise Words

Access the Delivered Noisewords page (Set Up CRM, Utilities, Search, Manage Noisewords, Delivered Noisewords).

The system displays all Verity-delivered noise words. Deselect or delete the ones that you don't want to use in verity search.

Maintaining Noise Words

Access the Maintain Noisewords page (Set Up CRM, Utilities, Search, Manage Noisewords, Maintain Noisewords).

Select the type of noise word list to view. If you select All, the grid displays the source of the entry.

Defining Record-Based Indexes

To define record-based indexes, use the Search Settings (RB_SRCH_PARMS) and Search Index Template (RB_SRCHIDX_TMPL) components.

This section discusses how to:

• Define settings for building search collection.

• Define record-based search index templates.

Pages Used to Define Record-Based Indexes

Defining Settings for Building Search Collection

Access the Search Settings page (Set Up CRM, Utilities, Search, Search Settings, Search Settings).

Maximum Number per Document	Specify the maximum number of permitted attachments per solution Images enbedded in solution details will not be counted toward the attachment limit.
Search Collection Name	Displays the name of the search collection.
Aging Interval in Days	Not used in PeopleSoft CRM searches.
Optimize Frequency	Enter the frequency for optimizing the collection. Optimization maximizes the size of the partitions that are in the collection. The default value is 5. This means that every fifth time you build the collection, it is optimized to reduce the number of partitions and maximize the amount of data that is stored in each partition.
Usage Increment and Usage Decrement	Not used by PeopleSoft CRM. Solution usage counts are based on actual associations with cases.
Documents per BIF File	Indicate the maximum allowable number of documents per XML/BIF file. The default value is 5000. Setting the maximum size of the XML input to the mkvdk Verity job with this limit assists performance when building a search collection.
Schedule Name and Job Name	Select the schedule and job that you established for this collection if you use a JobSet to build the collection. See Defining JobSets for the Build Collection Process. See PeopleTools 8.52: PeopleSoft Process Scheduler PeopleBook

Defining Record-Based Search Index Templates

Access the Search Index Template page (Set Up CRM, Utilities, Search, Search Index Template, Search Index Template).

Warning! PeopleSoft delivers one predefined search collection for Solutions (SEARCH_SOLUTIONS). Do not change this collection record index as there are codes tied to its record index (RC_SOLPRSRCH_VW). To format search results correctly, the system checks this record index. If you decide to change the record index of the Solutions search collection, you will also need to change the underlying code.

Search Collection Fields

The grid lists all of the fields in the record that are associated with this search index template. Do not remove any fields on the product catalog collection or the unstructured email collection.

Defining HTTP-Based Indexes

To define HTTP-based indexes, use the HTTP Index Definition (HTTPIDX_DEFN) component.

This section discusses how to:

• Define HTTP-based indexes.

• Select MIME (multipurpose internet mail extension) types.

• Select file names.

Pages Used to Define HTTP-Based Indexes

Defining HTTP-Based Indexes

Access the HTTP Index Definition page (Set Up CRM, Utilities, Search, HTTP Index Definition, HTTP Index Definition).

HTTP Index name

The name of the HTTP index. You can have more than one HTTP index.

Depths of links for crawling

Proxy Server

URL Options

URLs which contain the ? symbol typically lead to a CGI or other similar processing program. By default, these URLs are excluded from the build external content index process. Use this group box to enable indexing for URLs with the ? symbol.

Starting URLs

Selecting MIME Types

Access the Mime Types page (Set Up CRM, Utilities, Search, HTTP Index Definition, Mime Types).

Select to include specific MIME types. Any document type that is not selected will be ignored during the indexing process.

Selecting File Names

Access the File Names page (Set Up CRM, Utilities, Search, HTTP Index Definition, File Names).

Filenames

Defining File-Based Indexes

To define file-based indexes, use the File Index Definitions (RB_FILEIDX_DEFN) component.

This section discusses how to:

• Define file-based indexes.

• Manage file index filenames.

Pages Used to Define File-Based Indexes

Defining File-Based Indexes

Access the File Index Definition page (Set Up CRM, Utilities, Search, File Index Definitions, File Index Definition).

File System Path

Enter the file paths of file indexes to be searched at runtime.

Managing File Index Filenames

Access the File Index Filenames page (Set Up CRM, Utilities, Search, File Index Definitions, File Index File Names).

File Index Filenames

Configuring Search Options

To configure search options, use the Search Page Definition (RB_VRTY_PG_CFG) component.

This section provides overview on search behavior and word variations and discusses how to:

• Configure search definitions.

• Configure the record index list.

• View the HTTP index list.

• Define the file index list.

Understanding Search Behavior

This table describes each of the search behaviors that are available for searches:

Note. All search behaviors use the OR operator to process search requests. For example, if you enter the word computer for the with all words search behavior and you select both the Solution Description and Solution Summary fields as searchable fields, the system returns solutions that have the word computer found in their solution description or solution summary. Similarly, if the search behavior is without the words (and everything else remains the same), the system returns solutions that do not have the word computer in either the solution description or solution summary. In this case, if the word computer appears in the solution description but not the solution summary, the solution returns as a result of the search.

Understanding Word Variations

This table describes each of the word variations that are available for searches:

Pages Used to Configure Search Options

Configuring Search Definitions

Access the Search Definitions page (Set Up CRM, Utilities, Search, Search Page Definitions, Search Definitions).

Default Settings

Basic Search Options

The Basic Search Options group box sets the values of search options if the page searches in Basic mode.

Advanced Search Options

The Advanced Search Options group box sets the values of search option if the page searches in Advanced mode.

Configuring the Record Index List

Access the Record Index List page (Set Up CRM, Utilities, Search, Search Page Definitions, Record Index List).

One page can have more than one record index list and more than one record can be associated with a record index. The check box label corresponds to the record that is associated with the search index. Select check boxes that correspond to record zones within the record index.

Viewing the HTTP Index List

Access the HTTP Index List page (Set Up CRM, Utilities, Search, Search Page Definitions, HTTP Index List).

Select the HTTP spiders that you want to use for indexing web pages on a specific search page. The spiders that are configured for this page are the only ones available at runtime when search text is entered.

Defining the File Index List

Access the File Index List page (Set Up CRM, Utilities, Search, Search Page Definitions, File Index List).

Select the file index that you want the system to search at runtime. If the files are on an FTP server, select the FTP check box and an FTP server URL.

Building and Testing a Search Collection

This section discusses how to:

• Specify run control parameters and run the Build Collection process.

• Specify run control parameters and run the Build External Content Collection process.

• Test searches.

Pages Used to Build and Test a Search Collection

Specifying Run Control Parameters and Running the Build Collection Process

Access the Build Search Collection page (Set Up CRM, Utilities, Search, Build Search Collection, Build Search Collection).

Note. Normally, you can run a process immediately after defining the process parameters. However, if you use a JobSet, you must define the JobSet between setting up the run control and running the process.

See Defining JobSets for the Build Collection Process.

Run Control Details

Server Operating Environment

Search Collection Selection

Language Selection

Search Index Templates

This group box displays information about the search index templates that belong to this collection.

Additional Page Elements for Running the Process

Specifying Run Control Parameters and Running the Build External Content Collection Process

Access the Build external content index page (Set Up CRM, Utilities, Search, Build External Content, Build external content index).

To build a collection for external content, specify the external content type, the index name, the directory where the collection resides, and the languages for which the collection is built.

Testing Searches

Access the Query Test page (Set Up CRM, Utilities, Search, Query Test, Search Query Test).

Search Collection Name	Select the collection to test.
Parser	Select the parser to use for the query. Values are: Keyword Parsing, Free Text Parsing, and Query By Example Parsing.
Text to Search For	Enter sample text to search for. If you are testing the keyword parser, enter a keyword string into the field. If you are testing the free text parser or the query by example parser, enter a natural language string.
Search Using Verity Scoring	Click this button to view the results sorted according to the Verity internal scoring algorithms.
Search Using Usage Scoring	Click this button to view the results after they have been sorted using the usage count and last usage date and time calculations.

Defining JobSets for the Build Collection Process

To define JobSets and associate them with build collections:

1. Create run control records for the processes to run.

   Access the Build Search Collection page. For each collection that you maintain, create a run control record for each application server that is in the environment. Use a naming convention that identifies both the collection and the PeopleSoft Process Scheduler server name.

2. Create the jobs to include in the JobSet.

   Select PeopleTools, Process Scheduler, Jobs to access the Job Definitions page.

   1. Create a job that runs the Build Collection process once for each application server that is in the environment.

      For example, if you have four application servers in the environment, then the job needs to have four processes, each running the Build Collection process. Set the job run mode to parallel so that the job items can run concurrently on the different servers.

      PeopleSoft delivers a sample job, PRODCAT, that you can use as a model.

   2. If you are building the Product Catalog collection, create a job that runs the Load Catalog Cache process and the job that you defined in the preceding step.

      The Load Catalog Cache process loads the product catalog cache with the current contents of all catalogs. You must load the catalog cache before the collection is built, so put this process before the Build Collection process, and set the job run mode to serial.

      PeopleSoft delivers a sample job, BUILDCAT, that you can use as a model.

3. Create the JobSet.

   Select PeopleTools, Process Scheduler, Schedule Job Set Definition to access the JobSet Definitions page.

   Give the JobSet a meaningful name: If this is for the Product Catalog, then give it the name of the job that you created in the second step; otherwise, give it the name of the job that you created in the first step.

   The Run Control ID that you enter is overridden in the next step, so you can enter any value.

   If you want this JobSet to run automatically at regularly scheduled intervals, you can optionally attach a recurrence definition to this JobSet.

4. Set up JobSet items.

   Select PeopleTools, Process Scheduler, Schedule Job Set Definition to access the Schedule JobSet Items page.

   Add the jobs that you created in step 2 to the JobSet process list. Enter the following parameters for each:

   1. On the General Settings tab in the process list, enter the run control ID that is specific to a PeopleSoft Process Scheduler server.

      For example, if you have two PeopleSoft Process Scheduler servers named PSNT and PSNT2, and you're building the CRM_RB_SRCHDB collection (for solutions), then you might have run control IDs called SRCHDB_PSNT and SRCHDB_PSNT2.

   2. On the Server Settings tab in the process list, enter the PeopleSoft Process Scheduler server name in the Server Name field. Remember, each job runs on a separate PeopleSoft Process Scheduler server.

5. Associate the JobSet with the collection.

   Access the Search Settings page.

   Enter the schedule name and job name next to the search collection to which the search settings apply.

See Also

PeopleTools 8.52: PeopleSoft Process Scheduler PeopleBook

Using the FTP Option to Index Files

This section provides an overview on using the FTP option and discusses how to set up the FTP option to index files.

Understanding the FTP Option

When you set up the FTP option to index files, keep these considerations in mind:

• The filename of files on FTP server must not contain spaces.

• The application server and PeopleSoft Process Scheduler server must have access to the shared FTP drive and FTP path.

• Users will be accessing the files on the FTP server as anonymous user.

• The file index path name should be a substring of the 1st level of FTP server path and use the UNC path.

  For example, if the FTP server path is ftp://anonymous:anonymous@ADNTTP28/CRM, the shared FTP drive would be \\ADNTTP28\CRM.

• If the file index path (FTP path) has sub-folders, you must create additional URLs so the system will know which URL to use.

  For example, if the FILE_INDEX is already set to ftp://anonymous:anonymous@ADNTTP28/CRM and the CRM folder has sub-folders, you must create an additional URL. In this scenario, FILE_INDEX1 can be ftp://anonymous:anonymous@ADNTTP28/CRM/Folder2. If Folder2 has sub-folders, you must create another URL. In this scenario, FILE_INDEX2 can be ftp://anonymous:anonymous@ADNTTP28/CRM/Folder2/Level3.

• Make sure the URLs are set for the same file index definition in the Search page. In the Call Center Search page, add additional rows for the same File Index Definition but specify the different URLs on the File Index List page.

Setting Up the FTP Option to Index Files

Follow these procedures when setting up the FTP option to index files:

1. Go to PeopleTools, Utilities, Administration, URLs.

2. Create a new URL by entering ftp://anonymous:anonymous@yourservername/CRM in the URL field.

   Note. Depending on the depth of your FTP file path, create as many URLs as you need.

3. Go to Set Up CRM, Utilities, Search, File Index Definition.

4. Create a file index name. For example, CRM_FILE.

5. Enter \\yourservername\CRM in the File Path field.

6. Go to Set Up CRM, Utilities, Search, Search Page Definitions, File Index List.

7. Select the CRM_FILE file index name on the File Index List page.

Performing Search on the Universal Search Tool

This section discusses how to perform search on the universal search tool.

Page Used to Perform Search on the Universal Search Tool

Performing Search on the Universal Search Tool

Access the Universal Search Tool page (Set Up CRM, Utilities, Search, Universal Search Tool, Universal Search Tool).

Record	Select the record to search. Values include records for which you have defined search index templates. Selecting a value causes the Additional Search Information group box to appear.
Search	Enter record-level search text.
Display	Select the number of rows of data to appear in the Search Results grid. You can select 10, 25, 50, 75, 100, or 250.
Search Type	Select Phrases to perform a natural language search. Select Keywords to perform a Boolean search. If you select Keywords, click the Search Tips link for an explanation of the proper syntax for the search text. This setting applies only to record-level search criteria, not to field-level search criteria.
Match On	Select All Fields (AND) to search for data that meets all the field-level search requirements that you specify. Select Any Fields (OR) to search for data that meets any one of the field-level search requirements. This setting applies only to the field-level search criteria that you enter in the Additional Search Information group box, not to any record-level search criteria.
Search	Click to perform the search. The system searches the specified record for data that meets search criteria and then displays the matching rows of data in the search results grid.
Search Tips	Click to view tips for constructing keyword searches. For example, the search tips provide information on the use of and and or in searches and on searching for entire phrases.

Additional Search Information

This group box lists the fields in the record's search index so that you can enter field-specific search criteria.

Search Results

The following page elements appear regardless of the search record.

All other columns in the Search Results group box vary depending on the record that you searched and how you defined the search index template for that record. (On the Search Index Template page, select the Search Results check box to identify the fields that are to appear in the search results grid.) The columns display the data found by the search.