Poster, A Binary Decision Diagram based approach on improving Probabilistic Databases

A Binary Decision Diagram based approach on improving Probabilistic Databases

This research focusses on improving Binary Decision Diagram algorithms in the scope of probabilistic databases. The driving factor is to create probabilistic databases that scale, the rst order logic formulae generated when retrieving data from these databases create a bottleneck in the scalability in the number of random variables. It is believed that Binary Decision Diagrams are capable of manipulating these formulae in a more scalable way. This research studies the complexity of existing BDD algorithms with regards to characteristics such as the depth and breadth of a tree, the scale and other metrics. These results will be evaluated and compared to the characteristics of equations typically produced by probabilistic databases. We present an improved probabilistic BDD construction algorithm that scales better.

  • CS & BIT: Research Project

    The Research Project is a research project that serves as an exercise for the master’s thesis. As such it serves to give master students who did their bachelor study elsewhere the experience that bachelor students from the UT obtained during their bachelor project.

  • Research Paper

    View the full research paper for this project.

Poster, A Binary Decision Diagram based approach on improving Probabilistic Databases

A Binary Decision Diagram based approach on improving Probabilistic Databases

This research focusses on improving Binary Decision Diagram algorithms in the scope of probabilistic databases. The driving factor is to create probabilistic databases that scale, the rst order logic formulae generated when retrieving data from these databases create a bottleneck in the scalability in the number of random variables. It is believed that Binary Decision Diagrams are capable of manipulating these formulae in a more scalable way. This research studies the complexity of existing BDD algorithms with regards to characteristics such as the depth and breadth of a tree, the scale and other metrics. These results will be evaluated and compared to the characteristics of equations typically produced by probabilistic databases. We present an improved probabilistic BDD construction algorithm that scales better.

Kaj van Rijn

CS & BIT: Research Project

The Research Project is a research project that serves as an exercise for the master’s thesis. As such it serves to give master students who did their bachelor study elsewhere the experience that bachelor students from the UT obtained during their bachelor project.

Research Paper

View the full research paper for this project.