Summon: Including Your Library's Institutional Repository

Last updated
Save as PDF
Share
1. Share
2. Tweet
3. Share

Product: Summon

How do we include an institutional repository in the Summon service?

Use the following procedures to include an institutional repository (IR) in your Summon service.

When opening Support cases with requests regarding your local catalog (such as mapping changes, requests for new full load, questions about updates and deletes, requests for new institutional repository/Libguide, and so forth), please set the Category to Institutional repository/Summon local content.

You have the option to index one institutional repository in Summon when you first implement the Summon service; after that, you can add a maximum of one IR per quarter.

The repository must support the OAI-PMH protocol for Summon to automatically harvest the content. If the repository does not support the OAI-PMH protocol, then you will need to manually upload your files via FTP. After the records have been indexed for the first time, please use the notification form in the Content Ingestion section of the Summon Admin Console to notify us of any reloads, see instructions here.

To have your repository harvested, please submit a case using the Support Portal (available from the More Sites menu above) with the following options selected:

Asset: Central Discovery Index
Category: Client Content / Content Issue
Subject: Institutional Repositories (IR)
For the Title, enter one of the following:
- New IR -- [name of the repository] -- OAI-PMH (if the IR supports OAI-PMH)
- New IR -- [name of the repository] -- FTP (if the IR does not support OAI-PMH)

Include the following information in the Description text box:
- Name of the IR
- Public URL to the IR
- Expected record count
- URL for the OAI-PMH access point
  - Include specific instructions if necessary (e.g. sets)
- Should the IR be publicly accessible (to all Summon/Primo customers), or only accessible to your institution's patrons (that is, private)?
- Does the IR have any full text records, or is it abstracts/citations only?
  - Are there indicators in the records to identify full text vs citation?
  - Describe which records have full text; for example, "all dissertations are full text" or "all records with a format of 'application/pdf' are full text," etc.).
- How much of the IR content is Open Access (that is, the content is available to anyone to view)?
  - What is the Open Access indicator if the IR is a mix of Open Access and restricted content?
- Provide linking information so that all records have a direct link to the IR interface
  - Note: if direct links are not available in each record, we will default to the main homepage of the IR interface.
- Description of the repository content. For example, "Thousands of images, texts, and audio/video from ECU's diverse collections and beyond. Includes the more than 7,500 historic images from The Daily Reflector, Greenville, N.C.'s, daily newspaper, documenting the sweeping changes in Eastern North Carolina between 1949 and 1967."
- IR's Library Location (optional). Some libraries find that the Library Location facet assists users in accessing IR content. Note that one Library Location per IR is our standard practice.

Click Submit to send the request.

What happens next depends whether your IR supports OAI-PMH:

Repository Supports OAI-PMH Protocol

The Summon ingestion system will harvest all of the content from the OAI stream. After the OAI stream has been ingested, we will map the metadata and ingest the data into CDI.

We can use any metadata schema that your repository supports in OAI-PMH; however, we prefer to use OAI-DC, which results in faster ingestion. (See Mapping OAI-DC Fields for details.) Repositories using different schemas may take more time.

It will take approximately three to six weeks for the content to be processed and appear in Summon. After the initial ingestion, ProQuest will re-harvest the collection approximately once per week to ensure the content is up to date. If the content changes more or less frequently, please let us know when you submit the request.

Repository Submitted via FTP Transfer

After your request has been submitted, you will receive an email with instructions for uploading data from your IR to the Summon FTP server. The preferred upload formats are XML and MARC files. The data may also be submitted in tab-delimited text files with header-row field names.

Tab-delimited text files will take longer to upload and depend on the content format. Let us know if you plan to submit a tab-delimited file.

Once you have uploaded files to the FTP server, respond to the above email and let us know they are there.

The Summon metadata librarian team will map the metadata to the Summon metadata schema and then ingest the data into CDI. Allow three to six weeks for the content to be processed and appear in Summon.

To maintain the content in Summon, you will need to upload new, updated or deleted records to the Summon FTP server on a periodic basis, and let us know whenever you do so. If this is not done, the content in the Summon service will become out of sync with your repository. To let us know you have submitted a new full load for your repository, please use the notification form in the Content Ingestion section of the Summon Admin Console, see instructions here.

Basic Fields for Institutional Repositories in CDI

You can add Institutional Repositories to CDI in two ways, either public – where all institutions using CDI can benefit from the content - or private (as a catalog) for the contributing institution only. In order to qualify as a public collection, the data must meet minimum quality criteria, especially in terms of metadata depth. Records must contain title, publication date, ISSN / ISBN, DOI, Volume, Issue, start and end pages in separate fields as a minimum. Private collections do not have to meet such quality criteria and are subject to the institution’s own discretion.

Data Element	Requirement
Unique record identifier	Mandatory
Direct link to content or landing page	Mandatory
Title, subtitle	Mandatory
Open Access indicator – if not 100% at the collection level	Mandatory
Type of work (e.g. book, article, image) - if type is not indicated at the collection level	Mandatory
Abstract	Highly recommended
Author(s) / contributor(s) / editor(s) - full names where possible	Highly recommended
DOI	Highly recommended
Subjects / keywords	Highly recommended
Date of publication	Highly recommended (Mandatory in academic content)
ISSN / ISBN; eISSN / eISBN	Highly recommended (Mandatory in academic content)
Volume number	Highly recommended (Mandatory in academic content)
Issue number	Highly recommended (Mandatory in academic content)
Start page, End page	Highly recommended (Mandatory in academic content)
Journal title	Highly recommended (Mandatory in academic content)
Series name	Highly recommended
Alternative title (e.g. alternate language title, parallel title)	Recommended
Full text indicator - if not 100% at the collection level	Highly recommended
Grant - sponsor or funder ID	Highly recommended
Grant information (name or title)	Highly recommended
Language of original article	Highly recommended
Research IDs – ORCID, Scopus ID, ISNI, arXivID, ResearcherID, eRA ID	Highly recommended
Researcher affiliation	Highly recommended
Researcher email	Highly recommended
Re-use license information	Highly recommended
Thumbnail URI	Optional
Deletion indicator	Highly recommended

Article last edited: 19-Jul-2020

Date Created: 9-Feb-2014
Last Edited Date: 27-Apr-2016
Old Article Number: 8798