## High Level Framework of Big Data Graph Databases!

- 29/11/2016
- 194
- 0 Like

**Published In**

- Big Data
- Analytics
- Artificial Intelligence

In Big Data world, it was very much clear that the connected data to store and processing the data was first challenge. And the first ideation is to replace and leverage the tabular SQL Semantic with the graph-centric model.

And then the graph is new to big data world? Nope, the graph theory has been around for nearly 350+ years.

Currently the graph data bases are now used in industries as diverse as healthcare, retail, oil and gas, media, gaming, and beyond. And to take advantage of graph databases, we no need of deeper understanding of graph theory just understanding of graph is fair enough to leverage the power of graph and to get insights from the big data.

**So the second question, what is Graph?**

Graph is just a collection of vertices and edges and in other words it is set of nodes and the relationships. Graph represents entities as nodes and the way in which those entities relate to the world as relationships.

In big data era, Gartner identified five graphs in the world of business – social, intent, consumption, interest, and mobile; will provide sustainable competitive advantage if we leverages the above.

Below are given Graph Space Framework from 10, 000 feet high; which can be divided into to two parts – Graph database(technologies used primarily for transactional online graph persistence, typically accessed directly in real time from an big data app-OLTP), and Graph compute engine (technologies used primarily for offline graph analytics, typically performed as a series of batch steps-OLAP).

And other way to slice the graph space is to look at the graph models; the property graph (contains nodes and relationships, nodes contain properties (KV pairs), relationships are named and directed, and always have a start and end node, relationships can also contain properties), Description Framework (RDF) triples(it is a subject-predicate-object data structure, using the triples we can capture facts), and hypergraphs (it is a generalized graph model in which a relationship called hyper edge can connect any number of nodes.

**Graph Databases:**

It is an online database management with CRUD (Create, Read, Update, and Delete) methods to expose a graph data model. It’s specially built for transactional systems (OLTP). And we have to consider two properties while considering for graph database technologies – The underlying storage (native graph storage), the processing engine (index-free adjacency). Relationships are first-class citizens of the graph data model.

**Graph Compute Engine:**

It is a technology that enables global graph computational data algorithms to be run against large big data sets. It’s optimized for scanning and processing large amounts of information in batch. And some graph compute engine include graph storage layer.

Overview of graph data base space: Graph Storage in horizontal axis to the Graph processing in vertical axis.

**Conclusion:**

Graph Data Space is very powerful in big data era because it brings Performance, Flexibility, and Agility.

And the Graph Databases like Neo4j, FlockDB, HypergraphDB, AllegroGraph, Franz, InfiniteGraph are worth exploring during our big data analytics application design and architecture phase.

- 29/11/2016
- 194
- 0 Like

## High Level Framework of Big Data Graph Databases!

- 29/11/2016
- 194
- 0 Like

#### Kumar Chinnakali

Technical Architect - Insights and Data Practice at Capgemini

Opinions expressed by Grroups members are their own.

#### Top Authors

In Big Data world, it was very much clear that the connected data to store and processing the data was first challenge. And the first ideation is to replace and leverage the tabular SQL Semantic with the graph-centric model.

And then the graph is new to big data world? Nope, the graph theory has been around for nearly 350+ years.

Currently the graph data bases are now used in industries as diverse as healthcare, retail, oil and gas, media, gaming, and beyond. And to take advantage of graph databases, we no need of deeper understanding of graph theory just understanding of graph is fair enough to leverage the power of graph and to get insights from the big data.

**So the second question, what is Graph?**

Graph is just a collection of vertices and edges and in other words it is set of nodes and the relationships. Graph represents entities as nodes and the way in which those entities relate to the world as relationships.

In big data era, Gartner identified five graphs in the world of business – social, intent, consumption, interest, and mobile; will provide sustainable competitive advantage if we leverages the above.

Below are given Graph Space Framework from 10, 000 feet high; which can be divided into to two parts – Graph database(technologies used primarily for transactional online graph persistence, typically accessed directly in real time from an big data app-OLTP), and Graph compute engine (technologies used primarily for offline graph analytics, typically performed as a series of batch steps-OLAP).

And other way to slice the graph space is to look at the graph models; the property graph (contains nodes and relationships, nodes contain properties (KV pairs), relationships are named and directed, and always have a start and end node, relationships can also contain properties), Description Framework (RDF) triples(it is a subject-predicate-object data structure, using the triples we can capture facts), and hypergraphs (it is a generalized graph model in which a relationship called hyper edge can connect any number of nodes.

**Graph Databases:**

It is an online database management with CRUD (Create, Read, Update, and Delete) methods to expose a graph data model. It’s specially built for transactional systems (OLTP). And we have to consider two properties while considering for graph database technologies – The underlying storage (native graph storage), the processing engine (index-free adjacency). Relationships are first-class citizens of the graph data model.

**Graph Compute Engine:**

It is a technology that enables global graph computational data algorithms to be run against large big data sets. It’s optimized for scanning and processing large amounts of information in batch. And some graph compute engine include graph storage layer.

Overview of graph data base space: Graph Storage in horizontal axis to the Graph processing in vertical axis.

**Conclusion:**

Graph Data Space is very powerful in big data era because it brings Performance, Flexibility, and Agility.

And the Graph Databases like Neo4j, FlockDB, HypergraphDB, AllegroGraph, Franz, InfiniteGraph are worth exploring during our big data analytics application design and architecture phase.

- 29/11/2016
- 194
- 0 Like

## Kumar Chinnakali

Technical Architect - Insights and Data Practice at Capgemini

Opinions expressed by Grroups members are their own.