Browse Prior Art Database

Multiple Table Database Column Indexing Using Text Based Indexer

IP.com Disclosure Number: IPCOM000236402D
Publication Date: 2014-Apr-24
Document File: 3 page(s) / 37K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a method for indexing multiple tables within a single index, without the need to create a customizable solution. The structure and approach of this method can be applied to any Relational Database Management System (RDBMS) containing entity/attribute relationships.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 46% of the total text.

Page 01 of 3

Multiple Table Database Column Indexing Using Text Based Indexer

Many database indexing applications utilize text based search capabilities of Solr* (Lucene*) text based indexing or other text indexing. However, this does not enable distinguishing between different columns of an indexed table, since only column data is being indexed. In order to facilitate column specific indexing, a one-to-one mapping between table columns and indexed fields needs to be created, making every index a custom index servicing a specific set of tables.

A method disclosed herein provides a text based index capable of indexing multiple tables within a single index, without the need to create a custom solution. The disclosed method can be applied to any text-based indexer, including a Solr indexer. It does not prevent running traditional text based queries on the index, and can act as one if required.

Applications that use Solr as a text-based indexer lose the ability to run column-based queries. If column-based queries are required, a field for each column needs to be defined within an index schema, thus making the index very inflexible to be applied to multiple tables.

The disclosed method offers the ability to index a column or multiple columns from a table or a number of related tables in a database in a single or shared text-based index, such as a Solr text based index. . Depending on how the index is constructed, a query on an index returns a primary key of a row in a table or a primary key of a row in a parent table. Taking indexing out of the database dramatically increases data ingestion speed. Databases are built for data storage, not heavy data indexing; therefore, having a single index that utilizes a text-based indexing technology improves maintainability as

well as increases performance of data retrieval.

The method disclosed herein utilizes the same index for different database table structures, without any changes to an index schema, and still has the ability to run column-based searches on a text-based index, thus gaining speed and maintainability of a solution.

Solr requires a definition of a schema in order to create fields that can be indexed, stored, or indexed and stored. The method disclosed herein that allows multiple tables to be stored in a single index, as well as allowing index schema to provide fast column based searches that are similar to searching a database.

In an exemplary embodiment, assume a database table structure as described in Figure 1.

1


Page 02 of 3

Figure 1: Example database table structure

Assume that the database developer wants to index Race, Gender, Eye Color, and First

and Last Names, Year, Make, Model, and Id Type and Id Value. In order to accomplish this in a database, multiple indexes would have to be created, one for each column, or if the searches were done utilizing multiple columns, those combinations would also have to be created.

The same indexing can be accomplished much more easily using the di...