Welcome to this week in Neo4j, where we round up what’s been happening in the world of graph databases in the last seven days.
Featured Community Member: Alessandro Negro
I first came across Alessandro when he presented a talk on Reco4j, a recommendation framework that combined Neo4j and Hadoop, in June 2013.
He went on to create Reco4 before joining GraphAware.
Alessandro is currently leading the charge on machine learning/graphs and presented Mining and Searching Text With Graph Databases at our most popular London meetup event to date. In the talk, he showed how Neo4j and NLP can be combined to create advanced services on top of text analysis: recommendations, trend discovery, and finding influencers.
Alessandro will present Powering Relevant Search With Neo4j and Elastic Search at GraphConnect New York on October 24, 2017.
On behalf of the Neo4j community, thanks for all your work, Alessandro!
From GraphConnect: RDF vs. Property Graph Alternative Facts
We didn’t have an online meetup this week, so we’ll recap one of my favorite talks from GraphConnect Europe 2017 (Debunking Some RDF vs. Property Graph Alternative Facts) presented by my colleague Dr. Jesús Barrasa.
Jesus compares and contrasts the Labelled Property Graph (LPG) and RDF approaches to modelling graphs, the SPARQL and Cypher query languages, and addresses some common confusions when analysing these technologies.
You can also see the online meetup Graph Databases, RDF, and Linked Data, which covers similar ground.
Pentaho Kettle Plugin, Native vs. Non-Native, and Maven Dependency Graph
- know.bi's Bart Maertens announced the Pentaho Kettle plugin to load data into Neo4j.
- Dr. Jim Webber presented a webinar on Native vs. Non-Native Graphs where he explains the benefits of being able to represent data as a graph all the way down the stack from the query language to on-disk format.
- MZober1993 created profiling-tools, a tool that exports your Maven dependency graph into Neo4j and then allows you to query for dependencies and sub modules.
On the Podcast: Chuck Calio
This week on the Graphistania podcast, Rik interviews IBM’s Chuck Calio.
Chuck and Rik discuss running Neo4j on IBM Power8 and the benefits you get from running large graphs on that type of hardware.
Chuck also shares his vision for where he thinks graphs will play a role in the future of technology.
Chuck will be speaking at the GraphConnect New York conference on October 24, 2017.
Spring Data Neo4j, JVM Heap Analysis, and New Lynda Training Course
- Satish Peyyety created retailstore-neo4j, a set of RESTful services for an online retail store that runs on top of Neo4j using the Spring Data Neo4j library.
- Chris Leishman released v2.2.0 of neo4j-client, a command line shell for Neo4j. This version adds support for rendering byte array outputs.
- Nat Pryce and James Richardson‘s excellent Looking for Smoking Guns in a Haystack: JVM Heap Analysis With Neo4J presentation from 2014 resurfaced. In this talk, they show how to use Neo4j to track down a memory hogging JSON parser.
- My colleague Will Lyon released a new Lynda training course Database Clinic: Neo4j. Will introduces the Cypher query language, shows how to model data sets in a graph, and how to query them using the Python driver.
On StackOverflow: Promise.all(), Cartesian Products, APOC
This week on Neo4j StackOverflow…
- Tezraperformance tunes a graph global Cypher query by removing a cartesian product and brings down the execution time from 7700 seconds to 30 seconds.
- Bruno Peres suggests using the apoc.periodic.commit procedure to deal with a slow running import query.
Tweet of the Week
My favorite tweet this week was by Vova Kurbatov:
That’s all for this week. Have a great weekend!