Context
Release 1.8 of VIVO includes a search indexer that has been substantially re-written. Compared to earlier versions, it is more efficient, easier to understand, easier to configure, and easier to evaluate.
This increased level of visibility has made it obvious that the current configuration contains errors, redundancies, and omissions. It also brings this question to the forefront: what should the search index contain? How can this best be specified?
The restructured search indexer also lays the groundwork for added functionality. In particular, we would like to investigate the work done on faceted search results at the VIVO Hackathon 2014.
Objectives
To determine what data should be included in the standard distribution of the search index for Vitro. This should recognize that Vitro is ontology-neutral.
To determine what data should be included in the standard distribution of the search index for VIVO. This should be based on usage patterns and the structure of the VIVO ontology. It should recognize the tradeoffs between power and performance.
To develop a style for those configurations that is concise, easy to understand, easy to extend, and resistant to errors.
To make code enhancements that will support that style. This part of the effort should be minor.
To document the configuration process so individual VIVO sites can easily understand and customize VIVO search functionality to suit their needs.
To document the changes in content from previous releases.
To lay the ground work for configurable facets. To assess some use cases, potential roles for extensions like Blacklight, a preliminary design for the configuration file for the facets, affected classes, and jira tickets.
Deliverables:
A configuration for the search indexer, for use in Vitro.
A configuration for the search indexer, for use in VIVO.
A document or Wiki page that describes to VIVO maintainers how these configurations differ from those of previous releases.
One or more Wiki pages that describe the configuration process.
A requirements document for configurable faceted search, including use cases, an example configuration file, survey of affected code classes, and JIRA tickets.
A draft charter for a task force devoted to implementing faceted search.
JIRA ticket(s) for tasks that are raised by the task force, but are considered to be beyond the scope of the current effort.
Suggested schedule:
Develop charter | 1.5 weeks | 2015-02-06 |
Solicit members and schedule first meeting | 1 week | 2015-02-13 |
Develop action items, assign to team members. | 2 weeks | 2015-02-27 |
Accomplish deliverables, preliminary review | 3 weeks | 2015-03-20 |
Revise deliverables, final review | 2 weeks | 2015-04-03 |
Members
Jim Blake (Cornell University) - task force lead
Don Elsborg (University of Colorado, Boulder) - member
John Fereira (Cornell University) - member
Huda Khan (Cornell University) - member
Jing Wang (Johns Hopkins University) - member
Ted Lawless (Brown) - member
Meeting Times
TBD
Doodle poll for first meeting time is here: http://doodle.com/vrs8exmkz8ns92vm
Communication Channels
Distribute proposed charter and solicit participation from community via dev-all mailing list
Announce task force creation and progress on the weekly developer/implementation calls.
Create a page in the VIVO wiki (wiki.duraspace.org), with child pages for agenda, notes, and additional information
Utilize JIRA to assign, document, and track specific deliverables.
Agendas and Notes
- Draft agenda for first meeting is here: Search Indexer Task Force - First meeting