Leap Analysis

Machine Learning

NLP

Graph Abstraction for Databases

BUSINESS REQUIREMENT

The extraction and transformation of various types of data from different sources to a single format and database is a time consuming, expensive and redundant task in Data Analytics. Dexlock was keen to take up the challenge as an R&D project to result in performing data analytics tasks with more speed, security and efficiency.

CHALLENGES
  • The process of identifying the queries and converting the user defined scripts to the corresponding database queries for processing data.
  • The efficiency while converting these scripts into queries and running the queries in the respective databases.
SOLUTIONS
  • A semantic web analyzer was built to map data from different databases using a graph structure. The graph structure acts as a commonplace for queries to be performed.
  • An extension to an existing framework was custom designed to read data from different types of databases.
  • Custom graphs optimization tools were created to improve and monitor the overall performance of the platform.
  • The need of writing separate queries for each type of database was replaced with user-defined scripts. A custom solution designed to translate the user-defined scripts into the appropriate query type and format was implemented to achieve the same.
IMPACT
  • The platform overall is a major innovation in the field of Data Analytics.
  • The complex and time consuming process of ETL is simplified by this solution.
  • The redundant operations such as transformation of data into another format, separate query operations and the need for a separate database is no longer needed.
  • Data Integrity is maintained with the absence of moving data from the original location.
KEY TECHNOLOGIES
  • JAVA
  • ReactJS
  • D3JS
  • Apache TinkerPop
  • RDF
  • Spring Boot
  • Gremlin

More Projects

Category :