Research topic: "In-Memory Computing and Big Data Analysis with User-Defined Functions".

Research summary

When data started to grow, it’s impossible to stop growth process, and with such information expandability new challenges, like analytical scalability, must be faced. Within analytical scalability, migration of large dataset is a really a critical issue. Current existing techniques are mainly focused on performing data processing inside 3rd party solutions by manipulating exported (or moved, migrated) data. Such techniques are inefficient for large datasets. And as long as not only data exportation should be omitted, but also complex analytics provided, solutions for processing analytics inside database engine are becoming more challenging. Complexity, flexibility and efficiency of analysis can be hardly achieved only by SQL, and other approaches, like User-Defined Functions for DBMS, MapReduce paradigm, and In-Memory Computing, must be taken into consideration.

Main goals

  • Provide flexible and rich analytics for large datasets without data exportation;
  • Create framework (collection, library) for cross-DBMS analytics by utilization of User-Defined Functions;
  • Design methodology to support extension and enrichment of framework (collection or library) by adding new analytical algorithms and models;
  • Provide capability of designing new and tuning (configuring) existing analytical User-Defined Functions.


(Stand: 19.01.2024)  | 
Zum Seitananfang scrollen Scroll to the top of the page