Information about Curation Dashboard
Introduction
Curation Dashboard is a service originally developed by the technical team of the ACDH-CH hosted and maintained by CLARIN-ERIC. Its goal is to support CMD metadata authors and curators to improve quality of metadata for language resources. More information:
- CMD -Component Metadata - the CLARIN metadata framework.
- Code on Github
- CLARIN-PLUS deliverable D2.1 - specification document for the Curation Dashboard from 2016, formulated in the context of CLARIN-PLUS project.
Structure and functionality
Curation Dashboard consists of five subprojects:
- Curation-app
- Curation-web
- Link Checker
- Link Checker Persistence
- Link Checker API
Curation-app
This subproject does the actual analysis of individual CMD profiles, records and whole collections according to a number of quality criteria and generates reports and statistics which help discover potential problems that cause a lower metadata quality.
The reports are re-generated regularly (four-times per week) on the most recent dump of CMDI records as collected by the CLARIN-VLO harvester.
Curation-web
This is the user facing web application. It offers four main functions:
- on the fly validation of individual profiles and metadata records (either by their URLs or uploading them as files)
- presenting pre-computed statistics for CMD profiles
- presenting pre-computed statistics for collections
- presenting pre-computed statistics for link checking and a continuously generated statistics in detail view (hence values might differ if links of the provider have been checked in the meantime)
Link Checker
It checks constantly and repeatedly (with respecting the robots.txt files) all the URLs contained in metadata records of the collections. Checking means sending HEAD and/or GET requests to URLs and saving the results (just the request meta-information, the headers, not the payload itself) in a database. Core module later uses these results to generate statistics and takes them into consideration when assessing the quality of the metadata. Additionally, the results are also used by the CLARIN metadata catalogue, the VLO, to indicate the availability of a resource. This module is maintained in a separate code-base.
Link Checker Persistence
The Link Checker Persistence API (LPA) is a project, based on Spring data JPA, which serves as a common persistence layer for the other projects. This module is maintained in a separate code-base.
Link Checker API
The Link Checker API is a REST API based on SPRING BOOT, which serves as a web interface to the link checker database. It provides a mean to authorized users to upload their links to the Link Checker database for prioritized checking by the Link Checker. As well as questioning the Link Checker database current and historic status results. This module is maintained in a separate code base.
Frequently Asked Questions and Feedback
You may go to our faq page, which tries to answer the most common questions. If your question is not answered, feel free to mail us.