New Classes of Event Analytics

Event Analytics - With Snowplow, we need to empower our users to optimize their data. Where your data has major implications for the type of query and therefore your analysis can be enabled in it. Great time, we check data with SQL, and in particular, in Amazon Redshirt. This is a broad analysis of extensive OLAP styles - it allows us to slice and share different combinations of dimensions and metrics, for users, sessions, pages, and other entities we care about. However, when we do an event analysis, we often want to understand the sequence of events that occur. Maybe we want to know, for example:

How long will it take for users to switch from point A to B on our website or mobile app? What is the difference between the ways people go to get to point C? What are the different paths people take from point D? Event analytics pathing of this type is not supported with traditional SQL databases. You'll need to do a lot of table scans and run expensive windows functions for first-order events by users, then sort them. The graphics database is a new approach for obtaining and performing data queries. We've started experimenting with using to try some of the above questions. In this blog post, we will discuss the basics of graph databases, and start some experiments we have done with Neo4J in particular.

Event analytics often discussed with networks on social networks. They are used to modeling data where relationships are important, and Facebook, therefore, has a search tool called 'Graph Search'. The graphical database consists of nodes, which we can think of as objects, and the directed images, which connect nodes simultaneously. So on Facebook, your friends may be represented as nodes, their photos as nodes as well, and we can represent the love of one of their photos by creating an edge between the user node and the photo savings.