[Koha] Request for Support to Integrate ElasticSearch into Koha

[Sebastian Hierl]

Dear Colleagues,


 I recently sent this message to ByWater Solutions partner institutions,
but I would like to expand my call and solicit your support for integrating
ElasticSearch <http://www.elasticsearch.org/> into Koha.  The goal is to
add a new search engine next to the current Zebra search engine.  This is a
fundamental, strategic development that will ensure that Koha remains
relevant in the coming years and that will benefit the entire Koha
community.


 The current Zebra search engine works fine, but it is of a previous
generation and has important limitations.  In particular, it doesn't allow
us to take full advantage of linked data.  Koha currently links to Google
Books and Scholar, Amazon, the Open Library, and WorlCat.  This is
wonderful, but automatic links to Google Books and Amazon are limited to
recent publications with ISBN (in use only since the 1970s) and access to
Google Scholar, the Open Library and WorldCat is limited to a title keyword
search. With the proposed dependencies for implementing ElasticSearch
<http://www.elasticsearch.org/>, we will expand these options and
facilitate linking to most any repository by providing a MARC to RDF
crosswalk and greater flexibility to ingest other metadata schemes.


 Providing libraries with the option to select ElasticSearch
<http://www.elasticsearch.org/> as their preferred search engine instead of
Zebra (it is important to stress that this would be a choice), would open
the door to linking to the full content of Google Books and the Hathi
Trust, the Digital Public Library and other repositories, or to controlled
vocabularies, such as the Library of Congress Authorities
<http://id.loc.gov/> or VIAF <http://viaf.org/>.  Linking could be
established on a number of criteria, including keywords, but also OCLC
numbers and LCCN.


In short, implementing Koha with ElasticSearch
<http://www.elasticsearch.org/> would open the world of linked data and
permit Koha to function in a manner more similar to current discovery
tools.  Today, to benefit from the flexibility that discovery tools
provide, you would have to implement one in addition to Koha.  That means
contracting with another company or developing your own.  Few of us will be
able to afford that.  With ElasticSearch <http://www.elasticsearch.org/>,
the Koha community would be able to gradually develop Koha to fulfill the
same purpose, with the added benefit that the current development of
ElasticSearch <http://www.elasticsearch.org/> in Koha provides the option
to browse the collection.  In this manner, users would be able to perform
powerful and fast keyword searches, as well as browse the collection by
author, title, subject or call number, and link out to repositories for
access to the full text or to controlled vocabularies. Currently we cannot
browse the catalog in useful methods with Zebra.


 An additional benefit is that ElasticSearch <http://www.elasticsearch.org/>
will improve speed, as you can cluster the search engine and actually
control and configure the configuration files to use more system resources.
 The initial data is cross-walked from MARC to JSON, which is a lightweight
metadata schema and which will greatly improve speed.  Finally,
ElasticSearch <http://www.elasticsearch.org/> is open source and, just like
Koha, benefits from a dynamic community supporting its continued growth.


To keep Koha relevant and to expand its functionality mattersto all of us
and I would like to ask you to contribute to this development by contacting
brendan at bywatersolutions.com or kathryn at catalyst.net.nz -- please also feel
free to contact me.


Best regards from Rome,


Sebastian Hierl

*Koha Gruppo Italiano <https://it-it.facebook.com/KohaGruppoItaliano>*


--
Sebastian Hierl, Ph.D.
Drue Heinz Librarian, Arthur & Janet C. Ross Library
American Academy in Rome
Via Angelo Masina 5
00153 Rome
Italy

T: +39 06 5846 417
F: +39 06 5810 788