Resource Recommender for Primo VE

Introduction

The Resource Recommender service enables you to recommend the following resources based on searchable tags that you configure in Alma: database, website, person, library information, library guide.

When a user performs a search in which the query exactly matches the searchable tag (which may be a word or phrase), Primo VE displays up to three associated recommendations at the top of the brief results. Users can see all the recommendations by selecting the See all suggested resources link.

Resource Recommender Service on the Brief Results Page

Enhanced Search Capabilities

(New for November) Resource Recommender supports the following enrichments so that the search query does not have to exactly match the searchable tag to display the matching recommendations:

Type of Enrichment	Search Query	Searchable Tag
Synonyms	FDA	food and drug administration
Inflections (English only)	history	histories
Stop words (English only) are ignored in the search query and searchable tag: a, an, and, are, as, at, be, but, by, if, in, into, is, it, for, of, on, or, no, not, such, that, the, their, then, there, these, they, this, to, was, will, with.	harry potter stone	harry potter and the philosopher's stone By default, at least 75 percent of the search terms (not including stop words) must be included in the search query to return a recommendation. In this example, three out of four words match the tag.
Words ordered differently from searchable tag	harry potter	potter harry

Upon updating or adding new resource recommendations, tags are indexed immediately like other metadata changes to records.

This functionality is enabled by default. If you do not want to change the default setting of the minimum number of matching search terms, modify the value of the resource_recommender_tags_min_coverage parameter on the Discovery Customer Settings page (Configuration > Discovery > Other > Customer Settings). For more information, see Discovery Customer Settings.

Configuring the Resource Recommender Service

The Resource Recommender service enables you to list recommendations (such as contacts and databases) at the top of the brief results in Primo VE based on searchable tags that you define per recommendation. The Resource Recommender Configuration page lists all the recommendation types and enables you to manage each type's recommendations. Since only three recommendations appear in the brief results initially, priority is given to the recommendation types starting at the top of the list.

If you have configured more than 10K resource recommendations, you may experience performance issues while updating your recommendations. If this is an issue, it is recommended that you use the import/export methods to update recommendations.

To add a resource recommendation:

Open the Resource Recommender Configuration page (Configuration Menu > Discovery > Other > Resource Recommender Configuration).

Resource Recommender Configuration Page
Edit the following resource types to manage their recommendations as needed:
- Database – Displays the database recommendations on the left side of the All Suggested Resources page.
- Person – Displays the contact recommendations on the right side of the All Suggested Resources page.
- Website – Displays the Website recommendations on the left side of the All Suggested Resources page.
- Library Information – Displays the library information on the right side of the All Suggested Resources page.
- Library Guide – Displays the library guide recommendations on the right side of the All Suggested Resources page.
- Custom 1, 2, and 3 – Displays the custom recommendations on the left side of the All Suggested Resources page.
- Banner – Displays a recommendation that spans the entire page and appears above all other recommendation types, regardless of priority and persistence. The banner displays the following information: name, title, description, and link.
  
  Example Banner Recommendation
The Resource Configuration Page opens:

Resource Configuration Page for Suggested Database

On the Resource Configuration page for the selected resource type, select one of the following actions:

Add Resource (all resource types) - Enables you to add one or more resources manually.

In the Add Resource dialog box, specify the following fields:

Resource Recommender Fields
Field	Description
Key (required)	The case-sensitive key that uniquely identifies a recommendation and is used as the recommendation's code in the Resource Recommender Data Labels code table. When saved, the key is prefixed with the suggestion's resource type and separated with a dot character (for example, database.enteredKey) and cannot be edited. Per resource type (such as database), resources are sorted alphabetically by this field in the search results. If you intend to use the UResolver Electronic collection URL link constant for a suggested database, you must use the electronic collection's MMS ID as the key.
Name (required)	The name that displays for the resource in Primo VE search results.
Description	The resource's description (a maximum of 300 characters), that appears for the recommendation in Primo VE search results.
Main Title	The main title for the banner. This is applicable to Banner recommendations only.
Tags (required)	The search tags that trigger this resource recommendation. If you want to specify multiple tags for a recommendation, separate them with a semicolon. This field supports up to 3000 characters.
Link	The URL that is used to redirect users after they select the recommendation. Not applicable to Person resource types. In general, links that are not used for electronic collections to which your library subscribes must start with the following: http:// or https://. If you are adding database suggestions for electronic collections to which your library subscribes, you can specify the UResolver Electronic collection URL link constant if the suggestion's Key is the collection's MMS ID. This constant creates an OpenURL link to the database.
Link Text	If the Link field is specified, the system displays this text in Primo VE and redirects users to the specified URL when selected. Not applicable to Person resource types. This field supports up to 70 characters.
Image URL	The URL used to display an image for the recommendation in Primo VE. Not applicable to Library Guide and Library Information resource types.
Email	This field only displays for Person resource types. It provides the email address of the contact person.
Email Text	This field only displays for Person resource types. If specified, this text displays instead of the user's email address. If left empty, the Email address of the person is displayed. This field supports up to 70 characters.
Note	Optional descriptive text to explain the use of this resource to staff members who may be configuring recommendations.
Views	Select All if you want this recommendation to apply to all views. If you want to apply this recommendation to specific views, clear the All option and select the views to which you want this recommendation to apply.
Display Always	Select this option if you want this recommendation to appear for all searches. When this option is enabled, Tags are not required or used. Suggestions of type Banner have priority over all other recommendation types. The order in which other persistent recommendations appear in the suggested results is determined by the priority of the suggestion's resource type.

Select Add and Close to add the new recommendation.
Select Save to return to the Resource Recommender Configuration page.
Continue with the next step.

Add resources from electronic collections (Suggested Database only) - Enables you to add resources from electronic collections to which your library has a subscription. With this option, the system configures the following fields: Key, Name, Link, and Link Text. You must enter the tags before saving. The UResolver Electronic collection URL link constant creates an OpenURL link to the database. For more details, see the Key and Link fields.
1. Perform an electronic collection search of the repository and then select all the collections that you want to recommend.
  
  Repository Search for Electronic Collections
2. Select Select to save your selections and close the Repository Search dialog box. The selected electronic resources appear in the list of recommendations on the Resource Configuration page.
  
  Collections Added to List of Recommendations
3. On the Resource Configuration page, edit the configuration fields (such as Tags) as needed for each recommendation in the list and then select Save. For more details on each field, see Resource Recommender Fields.
  
  Edit Electronic Collection Recommendation
4. Select Save to save your changes and return to the Resource Configuration page.
5. Continue with the next step.

Enable each recommendation as needed.
Select Save to return to the Resource Recommender Configuration page.
Ensure that the resource types are activated.
The order in which resources display is determined by the order of the resources on the Resource Recommender Configuration page. Select the icon and move the row either up to give it higher priority or down to give it less priority.

Importing Recommendations

The Resource Configuration page enables you to import recommendations per the selected resource type using an Excel file. Unless noted in the following table, all columns must be included in the specified order. For example, if you are importing recommendations for Library Guides, you must omit the image_url , email, email_text columns from the import file. You may leave any non-required fields empty.

When importing more than 500 recommendations, imports are run in the background. When complete, an email is sent to the user. In addition, users can check the progress of the Resource Recommender Import job on the Monitor Jobs page in the Running or History tabs (Admin > Manage Jobs and Sets > Monitor Jobs).

Resource Recommender Fields
Order	Column	Description	Recommendation Type
1	active	Indicates whether this recommendation type is active or not. The valid values are true or false. If this column is not included or is left blank, the recommendation defaults to the active status.	All
2	key (required)	The case-sensitive key that uniquely identifies a recommendation and is used as the recommendation's code in the Resource Recommender Data Labels code table. For each key, you must add the *<recommendation_type>.* prefix. The following types are supported: banner, custom1, custom2, custom3, database, website, library_guide, library_information, and person. For example: library_information.library_hours Per resource type (such as database), resources are sorted alphabetically by this field in the search results. If you intend to use the UResolver Electronic collection URL link constant for a suggested database, you must use the electronic collection's MMS ID as the key.	All
3	name (required)	The name that displays for the resource in Primo VE. You must specify a display name for all recommendations and resource types. This field is translatable.	All
4	tags (required)	The search tags that trigger this resource recommendation. You must specify at least one tag per recommendation. If you want to specify multiple tags for a recommendation, separate them with a semicolon. This field supports up to 3000 characters. You can leave this field blank if the always_display field is set to true.	All
5	description	The resource's description (a maximum of 300 characters), that appears for the recommendation in Primo VE. This field is translatable.	All
6	main_title	The main title for the banner recommendation.	Banner only
7	image_url	The URL used to display an image for the recommendation in Primo VE. Remove this column when importing Library Guide and Library Information resource types.	Banner, Website, Database, Person, Custom1, Custom2, and Custom3
8	url	The URL that is used to redirect users after they select the display name in the recommendation. Remove this column when importing the Person resource type.	Banner, Library_Information, Website, Database, Library_Guide, Custom1, Custom2, and Custom3
9	url_text	If the url field is specified, the system displays this text in the new UI and redirects users to the specified URL when selected. Remove this column when importing the Person resource type.	Banner, Library_Information, Website, Database, Library_Guide, Custom1, Custom2, and Custom3
6	email	The email address of the contact person. Except for the Person resource type, remove this column.	Person only
7	email_text	If specified, this text displays instead of the user's email address. If left empty, the person's Email address is displayed. Except for the Person resource type, remove this column. This field supports up to 70 characters.	Person only
8	always_display	Indicates whether this recommendation appears for all searches. The valid values are true ( suggestion appears for all searches) or false (suggestion appears only when search terms match specified tags).	All
9	internal_note	Descriptive text to explain the use of this resource to staff members who may be configuring recommendations.	All
10	views (required)	Select All if you want this recommendation to apply to all views. If you want to apply this recommendation to specific views, clear the All option and list the views (separated by a comma) to which you want this recommendation to apply.	All

To import recommendations:

On the Resource Recommender Configuration page (Configuration Menu > Discovery > Other > Resource Recommender Configuration), edit the resource type to which you want to import recommendations.

Resource Configuration Page - Database
On the Resource Configuration page for the selected resource type, select Import.

Import Dialog Box for Resource Recommendations
In the Import dialog box, select the file to import (for example, Database_List.xls).

Example Excel Entry
Select one of the following buttons:
- Add Resources – Adds recommendations to the existing list of recommendations for this resource type.
- Replace Resources – Replace all recommendations for this resource type.
- Cancel – Closes the Import dialog box without importing.

Exporting Recommendations

The Resource Configuration page enables you to export recommendations per resource type to an Excel file. This enables you to modify many recommendations at once, create translations, and import the updated Excel file into the Resource Recommender page (see Importing Recommendations).

To export recommendations:

On the Resource Recommender Configuration page (Configuration Menu > Discovery > Other > Resource Recommender Configuration), edit the resource type from which you want to export recommendations.

Resource Configuration Page - Database
On the Resource Configuration page for the selected resource type, select the Export button and then select Excel.