graph database: All content tagged as graph database in NoSQL databases and polyglot persistence
James Kobielus summarizes Forrester’s predictions for 2012:
Enterprise Hadoop deployments will expand at a rapid clip.
In-memory analytics platforms will grow their footprint.
Assuming they are referring to products like SAP Hana, Tibco Spotfire BI, etc., my bet is that their adoption will depend heavily on their integration with Big Data toolkits.
Soon I also expect to see some in-memory data-grid products slightly shifting their direction and trying to penetrate the analytics market.
Graph databases will come into vogue: The market for graph databases will boom in 2012 as companies everywhere adopt them for social media analytics, marketing campaign optimization, and customer experience fine-tuning.
I know someone that will be very happy to read this prediction.
While I do agree this will happen, I also think that some more technical and communication advances in this space are needed before seeing a wide adoption of graph databases.
Original title and link: Forrester Predictions for 2012: Hadoop, In-Memory Analytics Platforms, Graph Databases ( ©myNoSQL)
Sir Tim Berners-Lee:
Inventing the World Wide Web involved my growing realization that there was a power in arranging ideas in an unconstrained, web-like way. And that awareness came to me through precisely that kind of process.
Let’s think how the different data models require us to arrange data:
- hierarchical model: free form, single-type of relationship (parent-child)
- relational model: strict form, (limited) multiple-types of relationships
- document model: free form, dual relationship types: logical and hierarchical
- star schema: strict form, (limited) multiple-types of relationships
Now think about graph databases: free form (nodes can have any number of properties), unlimited number of uni/bi-directional relationships. So question is, why aren’t network/graph databases used more these days?
Original title and link: Graph Databases and the World Wide Web ( ©myNoSQL)
I’m starting to catch up with the news after my sabatical month and it turns out things didn’t stay still during this period. While there are quite a few very important things that have happened during October, I’d like to bring up two very interesting ones that mark a possible turn in the NoSQL databases world.
The first insolvency/bankruptcy in the market.
This is an unfortunate validation of my thoughts about Graph Databases market penetration. sones GmbH has never been a market leader, but they could have tried to focus on a niche segment of the graph database emerging market and while that wouldn’t necessarily transform the company in a huge success, it would have probably gave it more time to refine the product and expand.
Update: Daniel Kirstenpfad (CTO, sones GmbH) reached out to me with some clarifications:
Achim Friedland was at a point in time the development lead of sones and in that position responsible for leading the developer team. He never was CTO of sones.
sones is not insolvent but rather is under preliminary bankrupty administration with the goal to arrive at a solution for continuation of product and company
I’m starting to notice a shift in the (marketing) message of a couple of NoSQL companies towards Enterprise NoSQL. I’m not yet sure what enterprise NoSQL means though: targeting enterprise customers, large scale NoSQL deployments, expensive NoSQL product and services packages, etc..
Whatever this terms means, I take it as a sign of: a) the market becoming too busy; b) growing competition for paying customers ; c) investors looking for clear validations of their investments.
What I hope this does not mean is the start of the unhealthy, unfriendly, and dirty competition. This market segment has greatly benefitted from a friendly environment in which all contenders have been pushing their products forward while working together to popularize and bring awareness to the polyglot persistence philosophy.
Original title and link: Two Important Events in the NoSQL World ( ©myNoSQL)
In case you were wondering how some problems Hadoop and MapReduce are not best at solving, there’s a great Q&A on Quora.com:
MapReduce is good at distributed computing, but not for graph algorithms. Is there a general-use, highly-distributed open source graph framework? I’m especially interested in hearing about in-practice use cases, and how good/bad they were.
Ankur Dave’s answer is quite compehensive, listing 5 specialized solutions and 3 generic frameworks:
- Golden Orb
I was not aware of all these solutions, so more to read for me.
Original title and link: What Are Some Good MapReduce Implementations for Graphs? ( ©myNoSQL)
Emil Eifrem (Neo4j) in an interview for StartUpBeat answering a question about the competition in the graph databases space (my emphasis):
There’s a lot of movement around alternative databases today and a lot of companies in NOSQL like MongoDB, Couchbase and Cassandra. However, when we’re out in the field and talking to customers, our actual competitors are in-house custom-built solutions.
This answer made me think that:
- Neo4j is the by far the graph databases market leader. And I’m not sure there’s a second place (InfiniteGraph maybe?).
- graph databases are still either unknown in many environments or perceived as niche solutions.
If I’d be a graph database producer, I’d not worry much about my product rank in the market. But I’d definitely be concerned about the current market size and graph databases market penetration in general.
Original title and link: Graph Databases Market Penetration ( ©myNoSQL)
Published by a group from Los Alamos National Lab (Hristo Djidjev, Gary Sandine, Curtis Storlie, Scott Vander Wiel):
We propose a method for analyzing traffic data in large computer networks such as big enterprise networks or the Internet. Our approach combines graph theoretical representation of the data and graph analysis with novel statistical methods for discovering pattern and timerelated anomalies. We model the traffic as a graph and use temporal characteristics of the data in order to decompose it into subgraphs corresponding to individual sessions, whose characteristics are then analyzed using statistical methods. The goal of that analysis is to discover patterns in the network traffic data that might indicate intrusion activity or other malicious behavior.
The embedded PDF and download link after the break.