Welcome to this week in Neo4j where we round up what’s been happening in the world of graph databases in the last 7 days.
This week we have a Reddit AMA on Russian Twitter Trolls, Machine Learning with Graph Databases, a Reddcoin Graph, and more.
The Neo4j GraphTour visits Berlin and London this week. Michael will be at both events and Ryan, our new team member Jennifer, and I will be at the London one. Don’t forget to register and come and say “Hi!” to us!
Featured Community Member: Tom Zeppenfeldt
This week’s featured community member is Tom Zeppenfeldt, Director & Founder at Neo4j Partner Graphileon, who build a development environment for interactive, graph-based dashboards.
Tom Zeppenfeldt – This Week’s Featured Community Member
Tom has been in the Neo4j community for several years and was one of the early interviewees on Rik Van Bruggen‘s Graphistania podcast.
He presented Interacting with Neo4j using a Graph-Driven User Interface at GraphConnect San Francisco 2016 and was interviewed by my colleague Bryce where he explained how the Graphileon tool works. Tom’s also been a regular presenter at meetups across Europe.
Tom regularly posts short videos and screenshots on his Twitter account showing how to use Graphileon to analyse different datasets so he’s well worth following on there.
On behalf of the Neo4j community, thanks for all your work Tom!
Reddit AMA: Russian Twitter Trolls
My colleagues Will Lyon and David Allen participated in a Reddit AMA about the Russian Twitter Trolls with NBC’s Ben Popken.
In the session Will and David explain their favourite findings, interesting Russian-isms in the data, and their hopes that other platforms will open up their data to allow similar analysis there as well.
Getting started with Neo4j, Neo4j-OGM, Spring Boot, Dr Aaron Clauset
- Jon MacKay has written a blog post in which he explains how to use a graph database to store and analyze complex networked information. Jon provides step by step instructions for getting started with Neo4j Desktop, formatting CSV files, and loading them into Neo4j.
- We have the 2nd part of Amy Hodler‘s interview with Dr. Aaron Clauset in which they look at how advancements in network science are impacting predictions based on connected data. They cover topics such as interpretability in deep learning models and building ensemble models to make predictions.
- Adam Cowley has written a post showing how to use Neo4j-OGM with Spring Boot. Neo4j OGM is a library that maps Java objects to nodes in Neo4j, similar to how ORM libraries are used when interacting with relational databases.
- Pol Dellaiera has a cool post in which he shows how to use Neo4j to solve a problem where you have to link a list of numbers such that adjacent numbers sums are perfect square numbers. You can see a video explaining the problem on the Numberphile YouTube account.
Pick of the week: An introduction to machine learning on graph databases
We’ve seen a lot of interest in graph based machine learning, so I was excited to come across a blog post written by David Mack in which he connects Neo4j with the Keras neural network library.
The challenge is to predict what review score a person will give to a product and David shows how to build a neural network based on the output of Cypher queries that achieves 100% test accuracy on this task.
Reddcoin Graph, Paradise Papers, Neo4j CSV Processor
- Johannes Unterstein has built a community graph out of the ReddCoin Twitter Tipbot usage. Reddcoin is a crypto currency that aims to enrich people’s social lives and make digital currency easy for the general public.
- In Enterprise Metadata Repository Facilitates Change at the University of Washington, Jennifer Zaino explains how they built a Knowledge Navigator using Neo4j to help with the migration of their HR system to the cloud. The Knowledge Navigator helped users work out where to find their data in the new system as well as providing a common vocabulary to help people navigate the data.
- In last week’s TWIN4j we mentioned the neo4j-csv-processor project, and this week Uwe Geercken has written a blog post showing how to use it on a sample dataset.
- I came across kutt, a Neo4j backed URL shortener that lets you set custom domains for your shortened URLs, manage your links and view the click rate statistics.
- Tomaz Bratanic has written a post in which he shows how to analyse the Paradise Papers using Neo4j’s graph algorithms library. Tomaz shows how to use PageRank, weakly connected components, betweenness centrality, and more to explore the sub graph of officers with a Swiss address.
Next Week
What’s happening next week in the world of graph databases?
Date | Title | Group | Speaker |
---|---|---|---|
February 26th 2018 |
|||
February 27th 2018 |
Mix of Neo4j and customer speakers |
||
February 27th 2018 |
|||
February 28th 2018 |
Graph-Powered Machine Learning and Chatbots with Neo4j and Amazon Alexa |
||
March 1st 2018 |
Mix of Neo4j and customer speakers |
Tweet of the Week
My favourite tweet this week was by Kenny Bastani, our former colleague in the Neo4j DevRel team:
Live coding with @Neo4j makes me look waaaay better at live coding than I actually am. Glad to make it a core part of my latest talk on @SpringData @springboot & @apachekafka.
— Kenny Bastani (@kennybastani) February 21, 2018
Don’t forget to RT if you liked it too.
That’s all for this week. Have a great weekend!
Cheers, Mark