June 2020 Summaries

This week, the Neo4j community has been busy with various content creators sharing their expertise. Will Lyon showed how to import spatial data for the GRANDstack Real Estate Search App, while Stefan Dreverman explained how to read and write data in the low-code platform Neo4j 4.1. Matt Holford and Ravi Anthapu gave a crash course on modeling patient journeys in healthcare using graphs. Ben Squire, a Senior Data Scientist at Meredith Corporation, was featured as this week's community member, sharing his experience with Graph Data Science. Additionally, Amr Khaled built an Instagram clone with the help of Neo4j and Cypher query language. The community has also been engaging on Twitter, with a tweet of the week from Roberto Rodriguez showcasing various projects related to Neo4j and cypher queries.

This week, the Neo4j community has been actively engaged with various projects and topics. Will Lyon started a series on building a GRANDstack Real Estate Search App, while Tomaz Bratanic explored the Node2vec embedding algorithm using Kaggle's Spoonacular Food Dataset. Meanwhile, Dan Flavin solved a triangle counting brain teaser, and Stefan Dreverman began a blog series on building a low-code platform with Neo4j. Mike Lam built a collaborative filtering-based recommendation engine based on beer data extracted from brewDB's Open API. The community also featured Stefan Dreverman as their member of the week, who is working as a freelance IT consultant in the Netherlands. Additionally, there were several other projects and topics covered, including a contact tracing testbed using Neo4j and Faker, an interview with Dr Alexander Jarasch from the German Center for Diabetes Research, and a tweet about AI accountability.

Graph technology is being used by Neo4j to accelerate research in medicine and life sciences, particularly during the COVID-19 pandemic. The graph database helps connect disparate datasets and parse textual data such as academic papers using Natural Language Processing (NLP). This technology is also being utilized by various organizations, including tellic in New York City, who are leveraging Neo4j's capabilities to analyze and connect large amounts of data. By utilizing graph technology, researchers can gain valuable insights and accelerate innovation in life sciences and healthcare.

The Graph Database community has been actively engaged in various projects and initiatives, with Karin Wolok introducing graph databases in a 10-minute video, Tomaz Bratanic exploring Lord of the Ring data on Neo4j, Chris Prenninger launching Neo4j Commander, a Graph App that simplifies updating graph data, and Nigel Small announcing the launch of version 1.8 of the Neo4j Go Driver. Meanwhile, Nathan Smith demonstrates how innovations spread across networks using Neo4j, while Sebastian Daschner, a featured community member, showcases his expertise on Neo4j and Kubernetes/Helm setup for Neo4j demoing it on the IBM IKS service. Additionally, Chris Prenninger launched Neo4j Commander, a Graph App that makes it easy to update graph data without having to write Cypher queries, and Nigel Small announced the launch of version 1.8 of the Neo4j Go Driver, which removes the dependency of the C connector library causing installation issues for users.

In a world with drastically evolving regulatory requirements, corporations are facing the challenge of adjusting their business processes and risk management practices. Anti-money laundering (AML) compliance is more than just ticking a box on a spreadsheet, as violations can be costly and result in significant damage to a company's reputation and brand. Effective compliance management requires a detailed "x-ray vision" beyond the first layer of huge data to understand, visualize, control, and mitigate risks. Graph technology is essential for creating an effective compliance management system, enabling companies to detect, analyze, quantify, document, and report suspicious connections and transactions between customers and business partner networks. A graph-based solution can help companies navigate complex regulatory requirements, manage risk, and protect their reputation.

The Neo4j Intelligent Recommendations Framework is built on a graph database that connects customers, products, services, vendors, and suppliers. The framework uses advanced technologies such as graph analytics and machine learning algorithms to produce highly relevant recommendations. A CPG company uses the framework to fill open positions with internal talent by analyzing employee job histories, skills, reviews, locations, and salaries in an extensive graph database. Each recommendation engine has a pipeline of Cypher queries called phases that build recommendation lists, including Discover Phase, Boost Phase, Exclude Phase, Diversify Phase, and others. These phases provide flexibility and power to deliver real-time recommendations, reduce query complexity, create easy-to-maintain engines quickly, score and weight phases depending on context, and enable building powerful engines for virtually any business case. The framework uses hybrid scoring, which permits the simultaneous use of multiple recommendation methods to select the best fit for each user and situation in real time.

This week's video features various graph-based problems and projects, including identifying graph-shaped problems, navigating the virus regulation pathway for COVID-19, building a movie recommendation engine, and exploring small-world phenomena. The Neo4j community has also announced new online training classes, with courses on Introduction to Neo4j 4.0, Graph Data Modeling for Neo4j, and Basic Neo4j 4.0 Administration. Featured community member Nathan Smith continues his series of blog posts about Networks, Crowd, and Markets, while Alicia Frame, Lead Product Manager of Graph Data Science at Neo4j, has an interview in Authority Magazine as part of the Female Disruptors series. The community also shared a tweet from Alexander Jarasch highlighting the impressive scale of the COVID-19 graph.

Graph Commons is an online platform that allows users to create graphs manually and visually, import data or interact with the graphs programmatically. It's a collaborative program that's free to use, targeting everyone from journalists to activists and artists. The platform has over 15,000 users worldwide in many languages, with thousands of graphs on display. Users can define nodes, connections, and properties, navigate through complex graphs, filter based on node properties, and apply simple analyses. Graph Commons is also a social platform where users can share their graphs online, embed them into articles or videos, and build communities around their graphs. It offers an API for structured data exchange and advanced calls using APOC. The platform provides options for selecting specific areas of the graph, exporting to PDF and SVG, and generating exhibitions. Real-life examples include Brexit analysis, Syrian refugees, artist collector networks, etymological maps, and hurricane impact on industries. Graph Commons is used by various organizations, including academic institutions, think tanks, and NGOs.

The text discusses a common situation in Neo4j, where there are large volumes of data across an entire customer base and only specific subgraphs of interest are relevant. To visualize this local part of the network, an analytical overlay is created on top of the graph to compute attributes such as total net trading volume from a given company and distance from A to the company in question. The overlay approach has several advantages over a simple approach, including not modifying the underlying graph, expressing metadata about the analysis, and easily pulling together the entire subgraph by traversing only one relationship. This allows for dynamic addition of different analyses in real-time and enables building standard BI visualizations on top of the data. The techniques described are beneficial for "schema-flexible" graph data modeling and Neo4j's relationship-oriented approach, making it easier to adapt these techniques to various problems.

The Neo4j Intelligent Recommendations Framework is designed to address the demands of modern recommendation engines, providing a scalable and adaptable solution for businesses. It takes advantage of Neo4j graph functionality to create a robust framework that supports various industries and use cases. The framework includes key components such as a graph database, connections to enterprise databases, hybrid-scoring recommendation engines, a Node.js server with Apollo GraphQL APIs, and a dashboard for managing the recommendations environment. By leveraging this framework, businesses can develop and maintain real-time recommendations solutions in a timely and cost-effective manner, meeting various business objectives while providing hyper-personalization for users.