August 2020 Summaries

The Neo4j reference implementation is a powerful tool for fighting money laundering by quickly identifying suspicious behavior and structures. By performing sprints with carefully chosen "money queries," developers can build simple solutions that reap the benefits of graph technology, such as localized pattern matching, graph algorithms, and entity resolution. These techniques use various methods like centrality, community detection, link prediction, similarity, and pathfinding to identify patterns in transactions and relationships between parties. The Neo4j AML graph data model is a whiteboard-style reference model that demonstrates best practices for working with the platform, requiring only twenty indexes to deliver millisecond response times at scale, whereas relational databases would require hundreds of indexes and much slower query-response times. By leveraging these techniques and tools, developers can quickly and efficiently identify money laundering patterns in their data.

In this week's video, David Meza presents a Knowledge Graph of Competency, Skills, and Knowledge at NASA, explaining how to build a graph that combines domain occupation ontologies with employees' individual skills. Max De Marzi starts a series of blog posts on using faux bitmap indexes to optimise single model queries in Neo4j. Will Lyon publishes Fullstack GraphQL Applications with GRANDstack, offering a free download of three chapters. David Makogon is featured as the community member of the week, having spent 30 years working with software and helping others through StackOverflow. The Summer of Nodes challenge concludes with exploring New York's Central Park graph. Max De Marzi also releases a blog post on using boolean filtering to improve query performance in Neo4j. Additionally, JQAssistant helps with refactoring and supports Neo4j and Docker, while Ng Wai Foong writes an introduction to the Neo4j Graph Platform for beginners.

Kineviz provides products and services for graph analytics that help business users gather, analyze, and visualize data to make informed decisions. Graph analytics enable the support of decision-making by providing insights into data relationships and connections. However, there are obstacles such as complex tools and interfaces that require technical expertise, which can hinder their adoption. The graph database enables assembling data from different sources in a coherent whole, while also being intuitive and corresponding closely to human understanding. Kineviz's GraphXR product supports business users by visualizing and filtering data, enabling discovery, aggregation transformations, and visualization for reporting purposes, ultimately facilitating the presentation of insights to stakeholders.

We have added a new course to our catalog of free online courses – Cypher Query Tuning 4.0, designed for experienced Cypher developers to learn best practices for writing performant queries. The course includes hands-on exercises and covers query tuning for graph data retrieval, as well as principles that can be applied to updating the graph. Key topics include understanding execution plans, reducing cardinality, optimizing property access, and monitoring queries. By completing this course, learners can gain skills to tune Cypher queries more efficiently, resulting in improved query performance.

Neo4j is a graph platform that stores data natively as a graph, allowing it to traverse paths with blazing speed and process complex patterns. This approach enables Neo4j to deliver fast query performance, in-memory analytics pipelines, and high-optimized data science libraries for detecting suspicious patterns and behaviors in real-time. With its end-to-end processing and retrieval capabilities, Neo4j is ideal for addressing the stringent requirements of anti-money laundering by providing blazingly fast query performance at extreme scale, causal clustering for disaster recovery, trusted integrity and dependability through ACID graph transactions, and enterprise-grade security through schema-based access control. Additionally, Neo4j offers a flexible and robust platform that extends to other global risk and compliance areas, with tools such as Cypher Workbench, Hop, and GRANDstack providing data modeling, input APIs, output APIs, analytics, and packaging capabilities.

This week's Neo4j community updates include a walk-through of the language drivers by Michael Hunger, announcements from David Allen and Elaine Rosenberg, and a REST API tutorial using django_neomodel by Sihem Bouhenniche. The featured community member is Nur Aini Rakhmawati, who has an impressive background in informatics and data analytics and will be speaking at Neo4j Online Developer Expo & Summit in October. The Summer of Nodes series continues with a murder mystery challenge, and Elaine Rosenberg launched an online training course on Cypher Query Tuning. Other updates include the launch of Neo4j AuraDB on Google Cloud Platform and releases from py2neo and Graph Gopher.

Many breakthroughs in sustainability are locked away in scientific journal articles, which is a problem for business developers who want to put sustainability research into action. However, using graph data science, specifically graph databases like Neo4j, can help analyze over a million journal articles and link science and business development. Planethon is working to bridge social ecological sustainability with the business world by connecting scientists and business developers through their consultancy services. They use Neo4j to scale up their research efforts and have started working with graph algorithms to enhance data analysis and find hidden information within publications. With Neo4j, Planethon aims to make sustainability research more accessible and scalable for businesses, ultimately helping them develop sustainable products and practices.

Neo4j has announced the general availability of Neo4j AuraDB on Google Cloud Platform (GCP), a fully managed graph database service that provides a seamless and simplified experience for cloud developers. The integration allows users to launch Neo4j AuraDB with just a few clicks, unify expenditures with their Google services, and automatically apply existing credits towards lower costs. With GCP and AuraDB combined, it becomes easy to use graph technology in any distributed cloud application, and users can spin up AuraDB to load data into a graph and start using Cypher and graph tooling immediately. The service is always-on, with database management on autopilot and no downtime or maintenance window. Users can scale the service on demand without interruption. Neo4j AuraDB is now listed on the GCP Marketplace, with a user-friendly control surface that integrates with Google Project billing. The pricing remains predictable, with pay-as-you-go pricing at an hourly rate, and costs are identical to when using the Neo4j.com experience. The current version of Neo4j AuraDB is suitable for small businesses and startups, but users can also opt for BYOL offerings on GCE and GKE or an enterprise Cloud Managed Service (CMS) offering.

Neo4j enables large banks to process millions of transactions daily while resolving entities and flagging suspicious activity in real-time, improving the accuracy and efficiency of anti-money laundering (AML) compliance. The graph database stores complex relationships between parties, institutions, behaviors, attributes, and times, allowing for more effective entity resolution and scoring. Neo4j's AML Framework reduces false positives and negatives, automates and instills rigor to the ER-scoring process, and cultivates domain expertise to model relationships. Graph algorithms help fill in gaps of missing and inaccurate data, identify relevant subgraphs, and combine normalized scores to produce an overall similarity score. The framework also enables investigators to visualize suspicious clusters and entities at a high level, providing more relevant facts and circumstances surrounding flagged transactions. By leveraging Neo4j's capabilities, compliance teams can better comply with AML regulations, make more accurate predictions, save money on regulatory fines, increase sales by improving brand reputation, reduce costs associated with false positives and false negatives, and meet stringent requirements for performance, availability, security, and agility at extreme scale.

This week's video features Will Lyon adding GraphQL authorization with Auth0 to the GRANDstack Real Estate Search App. The featured community member, Roni Das, shares his journey learning Neo4j and has been creating tutorials on Total Technology YouTube channel. Roni also offers consulting services around Neo4j, Python, and more. This week's video series includes Siddhartha Sehgal's graph modeling with Neo4j tutorial, Lju Lazarevic's Summer of Nodes challenge, Gaurav Sarma's explanation of Neo4j storage internals, and Mark Cleverley's Instagram graph building guide using py2neo library. Additionally, Arseny Zinchenko shows how to run Neo4j on Kubernetes, Victor Zhao explains SSL certificate generation for Neo4j, and Kashish Gakkar demonstrates connecting to Neo4j using the official Python driver.

This week, we are learning how to set up Neo4j Sandbox for Summer of Nodes. The Neo4j Sandbox is a collaborative and interactive environment that allows users to invite team members and stay live for up to 10 days. It's particularly suited for new users who want to play around with a simple example, enabling them to jump into the code and interact with different guides via desktop or console. For those who are especially new to Neo4j, Sandbox provides an immersive experience that can be accessed on-demand.

Graph-Powered Machine Learning is a book written by Dr. Alessandro Negro that explores the intersection of graphs and machine learning. Dr. Negro, a long-time member of the graph community, specializes in natural language processing, recommendation engines, and graph-aided search at GraphAware. He wrote the book to organize his ideas on this topic and provide a resource for newcomers to learn how to deal with complex tasks related to graphs and machine learning. The role of graphs in machine learning is changing rapidly, with graphs now being used for more advanced services and tasks. Dr. Negro expects that people will increasingly use graphs to solve their machine learning problems in the future. He has received positive early feedback on his book, which he hopes readers will not just read but also use as a practical resource. The book provides concrete use cases, datasets, and ideas that can be applied to real-world problems.

This week, the Neo4j community has been actively engaged with various projects and initiatives. Michael Hunger created a video exploring Spring Data Neo4j (SDN) 6, which will be part of the upcoming Spring Data Ockham release. The Summer of Nodes 2020 challenge was launched by Lju Lazarevic, where participants need to create graph models covering various aspects of barbecues around the world. Additionally, Michael Graham created a GraphQL API on top of Covid Graph, and Vidhya Murali et al. published CompoundDB4j, which integrates two prominent open source drug databases using Neo4j. The community also recognized Heinrich Klobuczek, who has made significant contributions to the Neo4j Ruby driver, as this week's featured community member. Throughout the week, community members shared their projects and experiences on social media, including a tweet by Uday Korlimarla praising Neo4j for being "fucking awesome" and addictive.

The GDS Library is a tool for data scientists that provides analytics-specific data structures optimized for global traversals and aggregation. It allows users to flexibly subset and reshape their graph for specific use cases, making it easier to work with large datasets. Alicia Frame and the developer will be presenting 7 tips on how to get more out of the GDS Library next week, which can be registered for now and used as a reference guide. The presentation aims to help users maximize the library's capabilities and improve their data analysis workflow.

The Neo4j community has been active, with a new video series introducing the low-code UI for loading RDF data into Neo4j, as well as a hands-on guide to building a mobile game and creating a Neo4j-backed application using TypeScript. The Developer guides have also been refreshed, with a new Graph Data Science section featuring videos explaining main concepts. A featured community member, Maxime Guery, has become an integral part of the Neo4j community, actively contributing and helping others through the Neo4j Ninja Program. Additionally, there are blog posts discussing Exploring DBpedia with neosemantics, Rock 'n' Roll Traffic Routing: Part 2, Getting started with Neo4j Fabric, Kong, New Elixir driver, Migrations in Kubernetes, and Exploring Graph Database Based Apps Using a Dynamic Model.