Page tree
Skip to end of metadata
Go to start of metadata

This work has transititioned into a Strategic Plan and Technical Roadmap

This "Product Plan" is more of an analysis of possible feature or use case gaps within the DSpace platform (based on the 2014 Vision Survey). It is NOT a true "Product Plan".
In addition, this document has been replace / reworked into a Strategic Plan for Technology and a corresponding Technical Roadmap. As of 2015, both are in progress and identified on the Strategic Planning page.
This analysis document has been left for historical reference, but please visit the Strategic Planning page for more recent strategic planning activities.

Candidate Features for DSpace

The following table represents a list of candidate features (and corresponding use cases) that DSpace software may wish to implement. They have been sorted by average ranking from the 2014 DSpace Vision Survey results (see Survey Average Importance), in order to show which features are of highest importance to our community. Each feature has been grouped into a rough "Category" (see Category descriptions), based on the part of the overall system it relates to. Finally, we've performed a rough analysis (based on corresponding use cases) of whether each feature has been implemented/achieved within the current DSpace platform (see Status explanations).

Overall, this table provides a way to visualize many of the highly ranked "use case or feature gaps" which have not yet been met within DSpace, and which likely should be prioritized for future software releases.

 

Candidate FeatureCategory*
(see categories below)
Status*
 (tick) = mostly met,
 (warning) = partially met,
 (error) = not met
Survey Average Importance*
10 = very, 
5 = moderately,
0 = not important

Use Cases

Create / manage files and metadata (as an Item)Structure(tick)10.00
Community and Collection hierarchy (or generic containers)Structure(tick) / (warning) (no generic containers)7.37

Create new versions of existing Items

Structure(warning) (disabled by default, limited)5.92
Support for derivative objects (e.g. thumbnails, coversheets)Structure(warning) (thumbnails & text extraction only)6.52
Metadata for all levels of object hierarchy (e.g. for Communities and Collections)Structure(error)6.99
Relationships between objects (e.g. Author items as Authors of an Item rather than textual metadata)Structure(error)8.17
Support for hierarchical metadata formats (e.g. METS / MODS)Structure(error)7.10
Item approval workflowsAdministrative(tick)7.88
Item embargo facilityEnd User / Administrative(tick)8.64
Support for flexible licensing, including Creative CommonsEnd User(tick)8.44
CRIS functionality
(Current Research Information System: researcher pages and automated publication feeds from external data providers)
End User(error) (though, DSpace-CRIS is available)5.51 
Search and browse for ItemsEnd User(tick)9.41

Easy and intuitive deposit mechanism for users

End User(warning) (arguable)8.86

Batch deposit (multiple item deposit via the user interface)

End User(error)8.59

Batch download (multiple item download as a single file)

End User(error)6.91 

Authentication with common single-sign-on systems (e.g. LDAP, Shibboleth)

Integrations(tick)8.52

Most configuration takes place via the administrative user interface

Administrative(error)7.61

Template-driven user interface for easy branding

Administrative(error)7.53

User interface and theme management takes place via user interface

Administrative(error)6.47

Search Engine crawler-friendly user interface

Integrations(tick)8.91

Specialized content delivery mechanisms (media players, page turners, document viewers)

End User(error)6.83 
Content streaming (video player)End User(error)6.92

 

Basic preservation (fixity checks)Administrative(tick)8.33
Format identification and reportingAdministrative(warning) (simplistic identification, no reporting)7.00
Advanced preservation services (e.g. format migration)Administrative(error)6.61
Persistent identifiers: handlesIntegrations(tick)8.32
Additional external persistent identifiers: DOIs, DataCiteIntegrations(warning) (basic support only)7.56
Other external identifiers: ORCIDIntegrations(error) (some current funding / development)7.22
Basic statistics (item access counts, file downloads)Stats/Metrics(warning) (limited reporting)9.03
Advanced statistics (search terms, geographical locations)Stats/Metrics(warning) (limited reporting)8.13
AltmetricsStats/Metrics(error)7.30 
Open Access status tracking and compliance checkingStats/Metrics /
Integrations 
(error)7.27 
Standard repository machine interfaces (e.g. OAI-PMH, SWORDv2, ResourceSync)Integrations(tick) (but, no ResourceSync yet)8.11
REST API for building external applicationsIntegrations(tick) (beta)7.56


Feature Categories

 Features have been grouped into rough "categories" as follows:

  • Structure: Features that relate to the architecture/structure of content within DSpace
  • Administrative: Features which are mostly administrative functions/activities
  • End User: Features relating to how end users interact with the system (deposit, download, search)
  • Integrations: Features which are either used by external systems (or programs) or interact themselves with an external system
  • Stats/Metrics: Features relating to metrics or statistics around content and its usage

Status explanation

(tick) - Arguably, DSpace meets most major use cases already. (However, there still may still be areas for improvement or enhancement.)

(warning) - DSpace meets some of the major use cases, but there are other major use cases which it does not yet meet.

(error) - DSpace does not really meet these use cases, yet.

Formula for "Survey Average Importance"

The "Survey Average Importance" represents average importance ranking of all respondents for a single feature from the 2013-14 DSpace Vision Survey (Survey results linked off of: DSpace 2013 Vision Discussions)

  • Very Important = 10 points
  • Moderately important = 5 points
  • Not important = 0 points

So, for example, "Community and Collection Hierarchy" had 50 respondents rank it "Very important" (10 points each), 37 rank it "Moderately important" (5 points each), and 6 rank it "Not important" (0 points).

This is an average score of ((10 x 50) + (5 x 37)) / (50 + 37 + 6) = 7.37 average (which is midway between a "moderately important" and "very important" ranking)

A spreadsheet was used to calculate all average scores. See the attached DSpace_Vision_Survey_Feature_Rankings.xls

 

Non-Functional Goals for DSpace

The following table describes high-level, non-functional goals which DSpace software should strive to achieve. These "non-functional" goals do not map directly to features or functions within DSpace, and in fact, some goals may be subjective in nature. However, they are perceived as highly important in order to continue to maintain and grow our active, open source community of users and developers.

Non-functional GoalUse Cases / Details
Easy to install / deploy
Easy to upgrade
  • "auto-upgrader" option? (script to upgrade, or it auto-upgrades data/configs when you load the new WAR, similar to Atlassian tools)
  • Store configuration properties/options in database
Scalability / Performance

Some Suggestions:

  • Uploading/Processing content in DSpace should not be bound to size of content
  • DSpace should strive to operate in with constant memory usage footprint during all operations.
  • DSpace user experience should not be performance limited by number of Communities, Collection, Items and/or Bitstreams
  • Support Alternative Assetstore Technologies
  • Support Many Database Vendors
Attractive to new developers
  • Opportunities to learn / contribute
  • Adopt Current and Industry Standard Technologies
  • Good developer and technical documentation
  • Supportive community
Attractive to new Repository/Content Managers
  • Good user documentation
  • Supportive community
  • Usage best practices / guidelines
  • More UI-Driven Admin Functionality
Avoid centrally maintaining duplicative codebases/functions
  • Standardize on a single out-of-the-box UI 
    • other third-party UIs may be offered, but would not be centrally maintained
  • Centralize common UI Actions / Commands and consolidate into Business Tier.
  • No labels