Graph Analysis & Special Techniques

Graphs are everywhere in business. Pretty much every abstract aspect of graphs, the many graph algorithms and mathematical theorems have direct applications in everyday life and in technology. In this page we have gathered the domains we are most familiar with.

We have done concrete projects in these domains in the past twenty years across various industries. Often the graph analytics aspect is part of big-data and machine learning but we welcome any type of business, whether you’re a startup or a large corp.

Books

Forensic Analysis

We help with the big picture and the small details, visualizing in various ways how large amounts of data create stories, how evidence creates proofs, how time and space joined with relational data create facts you can use in legal cases.

Graphs help complement the relational and tabular data, they allow centrality analysis and path finding, they allow to ‘see’ aspects hidden in the haystack.

Crime Prevention

How to see the narrative behind criminal cases? What is the story unfolding, what are the actors, the arena, the modus operandi?

We help combining data from different sources, visualizing and analyzing large graph repositories or help create them.

Using advanced machine learning, we help create inference engines and predict crime.

Epidemics

Epidemics can be seen as the transfer of information across networks and connected domains. The techniques to analyze epidemics apply well beyond the biological systems; the spreading of fake news, computer viruses, the adoption of fashions, cascading failures in electrical power grids, ecology of multiple species and so on.

We help companies analyze and define patterns, develop the appropriate tools and software and ‘see’ how the generic patterns evolve in their concrete business context.

Fraud Detection

We help defining and developing algorithms to detect fraud rings and anomalous behavior. Combining time series analysis and pattern detection in relational and/or geospatial data, we help you turning data into smart decisions.

Graphs are key when looking at organized crime and cross-usage of assets.


Books
Books

Knowledge Graphs

How to turn natural language into re-usable knowledge, how to store it and how to turn it back into language? How to organize domain-specific data and query it all?

Knowledge graphs and ontologies go together and there is a wide variety of options as can be seen from our graph database overview. We guide companies through the jungle of storage systems and software stacks.

Ontologies

Graph schema’s and ontologies are not very well known in many industries. Many NoSQL and graph solution even ignore it altogether. If you need a long-term knowledge management solution we help with creating ontologies and guide you through the many hurdles towards a knowledge base you can effectively use.

Whether you prefer triples or custom solutions like GRAKN.AI, TigerGraph or similar, we are here to help.

Marketing Optimization

Graphs are the secret ingredient when dealing with digital marketing optimization. Finding the hotpaths towards conversion, the transitions between touchpoints or modeling a Markov chain? We can help.

We helped very large corps applying graph analytic models to marketing challenges and seen big data in all its glory deliver its promise.

Social Network Analysis

Social network data has become the go-to data when augmenting in-house marketing data. We help companies with the creation of heterogenous networks and how to apply graph analytics to them.


Books
Books

Natural Language Understanding

Smart decision-making and (artificial) intelligence is never far away from NLP. We help companies applying deep learning to text and speech, how to convert language to graph structures (semantic labeling, question asking…) and how language in general can be used to augment existing data.

Counter Terrorism

Like crime investigation but on a grand scale, we help architecturing systems which embrace image, sound, language… in fact anything and everything to predict terrorism. This combines all the aforementioned techniques on this page and is often a well-kept intellectual property of companies and governmental organizations.