Oryzabase: Utilization of Solr to Provide Search Services

Oryzabase (www.shigen.nig.ac.jp/rice/oryzabaseV4/) is an integrated rice science database that is characterized by its original contents which are available publicly such as strain information (20,834 strains), gene information (8,461 genes) and journal information (8,605 articles). As a service to our users, we provide Rice ID checker, a search across species against 22 types of locus ID which are used in various databases. We also provide a Rice DNA Database which consists of Oryza information extracted from INSD data (International Nucleotide Sequence Database: DDBJ, ENA/EBI, NCBI). Previously, we used SQL queries and relational databases to implement these search functions. However, we faced problems with SQL queries not fully supporting orthographic variants and the search performance being remarkably slow with large amount of data.

In order to solve these problems, Oryzabase uses an open source full text search system called Solr. By implementing the advanced indexing features of Solr (support for inflected word forms, faceted search, auto suggest functionality), the search is more comprehensive, performs faster and the refine search is more advanced.

The search functionality of Reference Search, Rice ID Checker and INSD-Rice uses Solr and these services are available to public. Therefore, I would like to introduce more about the functionalities of these 3 systems.