By Neo4j Staff | February 26, 2014
To be honest, we were blown away.
When starting this challenge we were really excited and curious about the results. But what YOU created and submitted is just impressive.
We received 65 submissions in the 10+ categories. Well done!
Make sure to check them out, each one is a jewel on its own and there are many surprises hidden in these submissions. And if you get started with Neo4j one of these domains you might already have your modeling and use-case work halfway done. So before starting a proof of concept project, have a look.
You can certainly imagine, it was really hard for us to choose the winners (for sheer volume and quality). The quality of the submissions is really astonishing and we hope it will get even better when the feedback from the commenting sections is taken into account.
Everyone who participated will receive a Neo4j T-Shirt (if postal address and size was submitted here) and the winners will get an Amazon gift certificate (300,150,50 USD).
But without further ado let’s look at the categories and the winners:
- Interpreting Citation Patterns in Academic Publications: A research aid by Jonatan Jäderberg – an advanced use of graphs to connect scientific papers.
- Graphing our way through the ICIJ offshore jurisdiction data by @hermansm – an impressive investigative tracking of leaked data sets about (legal) company activities.
- Medicine & drugs classification for the Central Hospital of Asturias by @Roqueeeeee and @luigi9215 is an impressive representation of drug-related use-cases for a hospital.
- EPublishing: A graphical approach to digital publications by @deepeshk79 impressively covering a lot of different use-cases in the publication domain and workflow.
- QLAMRE: Quick Look at Mainstream Renewables Energies by @Sergio_Gijon is a quick look at categorizations of renewable energies.
The Antarctic Research: The Effect of Funding & Social Connections in the US Antarctic Program by @openantarctica is really impressive but sadly not eligible as the demo dataset used is too large for the limited scope.
Transport and routing is a great domain for graphs and we see a lot of potential here, unfortunately the sandbox is not well suited for the some of the large demo datasets, so some of the entries did not qualify.
Advanced Graph Gists
As expected this has been most impressive, people really went far and wide to show what’s possible with graphs and graph-gists. Really hard to choose in this category.
- Movie Recommendations with k-NN and Cosine Similarity by @_nicolemargaret Nicole really shows off, computing, storing and using similarities between people for movie rating.
- Small Social Networking Website by @RaulEstrada this is not over the top like others but a really good and comprehensive example on what graphs are good for.
This category unintentionally sneaked in but had some really good submissions. So we also award some prizes here. It’s like the little brother of the Advanced category.
- Embedded Metamodel Subgraphs in the FactMiners Social-Game Ecosystem Part 2 by @Jim_Salmons explores the possibilities of using data and meta-data in the same graph structure and which additional information you can infer about your data.
- User, Functions, Applications, or “Slicing onion with an axe” by @karol_brejna covers resource and permission management of an IT infrastructure.
I not only want to thank all of you who contributed, but also our awesome judging team (Mark, Wes, Luanne, Jim, Kenny, Anders, Chris) who spent a lot of time looking at the individual GraphGists and provided valuable feedback in the comment sections. So please authors thank them by updating your gists and taking those comments into account!
As we want you to always publish your awesome graph models, we’d like you to know:
Everyone who, now or in the future, submits a GraphGist on a new topic
via this form will get a t-shirt from us.
GraphGists are a great initial graph model for anyone starting with Graphs and Neo4j.
That’s why we want you to vote on the gists your really like or found helpful.
In case you wonder what the “Rules for a good GraphGist” are, that we used for judging, here are some of them. So if you work on a GraphGist in the future, please keep them in mind:
- interesting/insightful domain
- a good number of realistic use-cases with sensible result output
- description, model picture should be easy to understand
- sensible dataset size (at most 150 nodes 300 rels)
- good use of the GraphGist tools (table, graph, hide-setup etc)
- we had an epiphany while looking at the gist
And last but not least a special treat. The structr team has added GraphGist import to structr so you can automatically create a schema and import the initial dataset into your graph-based application. Then add some use-case endpoints and you’re done.
Michael for the Neo4j Team