Xomnia is a word combining the letter ‘X’ – the unknown – and “Omnia” – Latin for everything. Our team of data scientists and big data engineers are trained to find the undefined – X – in all the relevant data sources – Omnia. This unknown – X – is untapped business value. Combining the X and Omnia you get the Xomnia spirit. Eager, curious and dedicated people, who have the belief that the future is big data.

Want to feel the Xomnia spirit? Follow us

News

XomniaNewsShort on data? Get a new perspective with graph analytics

Short on data? Get a new perspective with graph analytics

More often than not, data scientists work with huge amounts of data that are produced by a business. They typically use numerous mathematical, statistical and programmatic techniques to extract valuable information and get untapped value out of data.

As a data scientist, you might face a challenge at some point of an insufficient amount of historical and/or labeled data. As a result, the majority of knowledge discovery methods can’t be used effectively. Without being able to produce extra information for business inquiries, we tend to move on from the project. However, I’d like to challenge you to look at your data from a different point of view. I can show you the value of graphs by showcasing examples in anomaly detection applications where one is interested in finding the most unusual data occurrences.

A great deal of information can be discovered not only from data records themselves, but also from analysing the relationship between the data points.  Analysts can begin utilizing more out of what they already have by trying to understand the relationship between data. So-called graphs and graph analysis (a.k.a social networks or networks) can help you to discover different interactions, communications and relationships between one, or multiple data points. This enables you to imply more measures and context onto the data.

What is graph?

Graph (a.k.a social network) is an ordered (for directional graph) or unordered (for undirected graphs) pair of vertices and edges (a.k.a. nodes and links), where vertices can be represented as a circle, and edges as a line connecting circles. The most common graph examples are family trees, tournament brackets, organisational charts, IOT charts, or Facebook and Linkedin friends/colleague networks. Most commonly, graphs are interpreted as an adjacency matrix. This means that relationships can be processed by any conventional programming language and stored in any database.

Graph data structure represents each data point as an individual instance. In comparison to Entity Relation (ER), the data structure relationship between individual nodes can be unique. Graphs can be seen as a different paradigm to the traditional ER representation. Yet, it also comes with its own pros and cons.