Back to the Future of your Data - Wrangling connected data over time
A Talk by Dexter Lowe (Principal Engineer, G-Research)
Login To Watch This Content
LoginAbout this Talk
This presentation will delve into the intricate practice of modelling the evolution of a graph over time, as well as our changing view of the timeline itself. This knowledge will empower you to comprehend and interrogate your connected data over time, providing a deeper understanding of the data's development from many temporal viewpoints.
This presentation represents the culmination of more than a year's worth of experimentation with temporal modelling in graph databases, with some of the best minds in this field. The solutions discussed have been applied to effectively model a dataset comprising billions of nodes, in a manner that enables us to observe any historical state and readily identify when changes occurred.
What will I learn:
The presentation will encompass abstract discussions as well as practical examples in various query languages. A general understanding of property graphs and fundamental data querying techniques will enhance the experience. Nonetheless, the talk is designed to be approachable and enlightening for a broad audience. Familiarity with modelling time in graphs is not necessary.
We will cover several solutions to timeline modelling in graphs such as the Entity-State model and Discrete Time Dynamic Graphs, their limitations, as well as how to model additional time dimensions using these approaches, and query them effectively.
Attendees of this talk will come away with a solid understanding of:
* Why you might want to build extra time dimensions into your data model
* How to deal with the evolving net of connections over time
* How to handle multiple timelines
* Scalability considerations
* My favourite tea for a rainy afternoon.