DTU Findit is the Library Services Platform (LSP) of DTU and is technically operated locally by DTU Library, and consist of a number of technical components, most of which are also developed and maintained inhouse by DTU Library and based on Open Source technologies, supplemented by some community/commercially developed.
Components include:
- Data Acquisition, Processing and Indexing (Processing Pipeline):
- Acquisition & processing – taylormade platform, optimized for data from data providers in the STEM area and relevant to DTU (perl, postgresql, SOLR)
- Merge & index – taylormade platform, optimized for indexing and query performance required by DTU patrons (java, SOLR Cloud, ZooKeeper, postgresql)
- Configuration & flow control – taylormande, optimized for DTU’s OpenScience engagement by ensuring currency (minimal delay of daily data updates) while at the same time delivering data quality enhancements for DTU patrons by allowing for resource consuming data reprocessing, taking advantage of data retransmitted by providers when they – pushed by the Open Science agenda - enhance quality and integrity of their data (RabbitMQ, postgresql, ActiveAdmin)
- User interface:
- Taylormade discovery & delivery layer, optimized for DTU patrons research and study related information needs and behavior (Ruby on Rails, postgresql)
- Access: Through the User Interface, but in practice handled in the background by various microservices:
- Direct Access – linking directly to Open Access publications or through link resolution (SFX by ProQuest/ExLibris) and active proxying (EZProxy by OCLC) for paywalled but licensed publications, optimized for frictionless, location-independent access
- Access by Request/ILL - through a taylormade publication request handling platform, optimised to use international document suppliers combined cost-efficiently with local librarian based request/ILL services (java, postgresql, redmine)
- Print Book lending – through a Catalog & Circulation system optimized for DTU’s book collections and lending policies/practices managed by DTU Library (koha)
- Print Journal scans – through a taylormade registry of DTU’s static print journal collection, optimized for DTU’s scan-on-demand service, managed by DTU Library (Ruby on Rails, ActiveAdmin)
- Integrations: Taylormade integrations to various DTU services and systems, including
- DTU person directory
- DTU research output registration
- DTU Learning management
- DTU annual reporting (UMV) on research output
For efficient uninterrupted inhouse development, maintenance and operation of the taylormade components, focus is on stability (debian), automation and containerization (docker, kubernetes, ansible) aa well as continuous integration (gitlab).
For staying relevant to DTU patrons, project management is anchored at DTU Library management and conducted using efficient best practice methodology (iterative, agile development) and project tools (redmine, mattermost), and focus is on activating best practice usability tools (including UXlib) in order to capture and understand user needs as accurately as possible.
DTU Findit comply to GDPR and DTU IT security paradigm.
Developing, maintaining and operating locally inhouse an LSP based primarily on Open Source technologies has been done by DTU Library since the mid 1990’s – long before the term LSP was coined. DTU Findit came about in 2013 and is the latest generation. Previous generations include “DTU Digital Library” and “DADS”.
