Publication Date: 2010-Jun-22
The IP.com Prior Art Database
By polarizing the data storage into two tables, during search or query, the volume of target data gets reduced, ideally by 50% in each of the tables. From the keyword given by the user in search functionality or from the condition-value (literal or variable value) present in search queries, it is possible to decide on which polarity table to target the query against. Additionally, due to reduced volume of targeted data, performance of searches and queries are enhanced. Further, as the CPU is less engaged, the system is more scalable.
FIELD OF INVENTION
The invention generally relates to database searching methods and more particularly to a technique which increases the overall performance and scalability of a database.
BACKGROUND OF THE INVENTION
In a generic search performed using Google web page, the entire Web is searched for a given keyword for its presence anywhere in the trillions of words stored in the Internet. To enhance the response time, several techniques such as Indexing, Depth-First and Breath-First among others are used. These techniques are relevant only to the Web context.
Generally, in structured data storages such as relational database management system (RDBMS) databases, data is classified into defined schemas. Due to the classification of data into schemas, searches for a keyword are targeted to a specific column/field (target column-TC) of a table (target table - TT) in a database (target database-TB). In the initial stages of software implementation in a productive system, due to low volumes of data storage in the database, performance is not a major concern. However, when sufficient attention is not applied to the software logic foreseeing the growth of data, performance in terms of rendering the data to the screen, especially during search queries is greatly affected. The affected performance results in customer complaints on the software's non- usability in a production environment leading to customer dissatisfaction and loss of goodwill.
Hence there exists a need for a technique to increase the overall performance and scalability of simple search queries in structured data storages.
BRIEF DESCRIPTION OF THE INVENTION
A polarized database which enhances performance by polarizing various data types to arrive at data polarity is disclosed. Further, the data polarity of columns is used to arrive at row polarity and the transaction rows are segregated into two or multiple tables based on the row polarity.
DETAILED DESCRIPTION OF THE INVENTION
Generally, in structured databases such as RDBMS, schemas define the structure of the data that is stored. In schema-based data, at a high level, the attributes are categorized in two ways as key data and column/field data. Key data constitutes primary and foreign keys that enable random access & referential integrity. Column/field data comprise data for the attribute that constitutes the profile of the entity. Further, data is typically classified into but not limited to certain categories of data types. These data types comprise binary valued data, numeric valued data, date valued data and character/text valued data.
The present invention provides a strategy for storing data. Each of the above mentioned data types of a typical database is classified into either bipolar or multi-polar categories such as, positive/negative data, odd/even data among others. Classifying data into bipolar or multi-polar categories allows the data storage to be accordingly segregated, resu...