January 2021 Summaries

Neo4j AuraDB Enterprise has been launched, targeting large organizations building global-scale graph applications. Neo4j AuraDB FREE is also being developed and will be released soon. A video from Emil Eifrem's presentation at the Open Core Summit is available. New features in Bloom 1.5 include a mini map for simplified graph navigation and performance enhancements. HDBSCAN Clustering has been implemented using the GDS Library and APOC. Neo4j AuraDB team member Irfan Nuri Karaca has been featured as this week's community member, with a long history of contributions to Neo4j projects. Additionally, new tools and features have been released or announced, including Neogma Nodes.js OGM and Gram data graph format.

BearingPoint's business technology and analytics teams work closely together to guide clients through the full process of identifying business pain points, designing and implementing full-stack graph solutions. Graph databases provide a new way of storing data entities and their relationships, offering benefits such as better insights into connected data, scalability, and ease of retrieving data. BearingPoint has strong expertise in both business and technological aspects, allowing them to organize implementations and create working software. They have guided multiple clients with successful graph database implementations, including a knowledge institute and an HR-focused resourcing graph, which provide valuable insights for personalized search results and employee resource planning. The team starts by identifying client problems, aligns on expectations, designs use cases, and connects data sources to leverage technology and analytics expertise.

Neo4j AuraDB Enterprise is a fully managed cloud database service designed for large organizations building mission-critical, global-scale graph applications. It offers performance at scale, enterprise-grade security, predictable pricing, reliability for essential applications, and 24/7 premium support. The service is now available on Google Cloud and Amazon Web Services (AWS) as part of an early access program, bringing Neo4j closer to a multi-cloud future. Neo4j AuraDB Enterprise uses Kubernetes container orchestration and Neo4j's Causal Clustering technology for self-monitoring and self-healing architecture, ensuring the reliability and consistency of graph data. The service also inherits security and resilience characteristics from public cloud providers' services, including end-to-end encryption and secure cryptographic protocols. A highly skilled team of engineers maintains and runs Neo4j AuraDB as a production system, sharing operational responsibilities across the product engineering team to ensure reliability and continuous learning.

NeoDash is a tool that allows users to build Neo4j dashboards without writing any front-end code. It connects to the user's Neo4j database and runs Cypher queries to populate reports, which can be customized with various parameters. The current version of NeoDash supports five types of data reports, including graph visualizations, table views, bar charts, line charts, and JSON output. Users can create dashboards by specifying a single Cypher query for each report, and customize the appearance of the reports using settings menus. Dashboards are automatically cached in Neo4j Desktop or web browsers, and users can export them as JSON to share externally. The tool also includes features such as hard resets and customizable parameters, allowing users to fine-tune their dashboards.

Graph databases are being used by companies to solve complex problems, such as real-time recommendations. These databases can combine a customer's browsing behavior and demographics with their buying history to provide relevant recommendations. Graph technology is particularly useful for connecting masses of buyer and product data, allowing for the analysis of relationships between entities and understanding the quality and strength of those connections. In particular, eBay uses graph technology, powered by Neo4j, to generate real-time recommendations that take into account contextual information and natural language understanding. This approach enables the creation of a real-time recommendation engine that can understand and learn from shopper interactions, providing personalized product suggestions. The use of graph databases in this way is becoming increasingly important as consumer expectations for relevant and timely suggestions continue to grow.

Gram is a textual format for data graphs that aims to provide a common language for working with graph technology, including databases, analytics libraries, visualization software, and other tools. It was designed to complement the Property Graph Model defined by ISO GQL (Graph Query Language) and addresses scenarios such as describing relationships between entities, sharing data while reserving space for individual use, and representing complex information structures like paths and associations. Gram is friendly to read and write, composable, sliceable, pageable, and streamable, and vendor-neutral. It uses a path-based representation inspired by Cypher path expressions and provides a way to represent data graphs in a format that can be easily understood and manipulated by humans.

The Developer Relations team at Neo4j has been busy sharing various content, including a video introduction to Liquigraph, a database refactoring automation tool for Neo4j, and blog posts on using QuickGraph with a SKOS taxonomy, the Neo4j Traversal API, and building an asset management system. The team also highlighted the work of community members, such as Dominic Kumar, who has been actively contributing to the Neo4j community since 2018. Additionally, there are updates on Obsidian Neo4j Plugin, Neo4j Traversal API, and GraphAware UUID on Neo4j 4.x, including a paper on using graphs for explainable decision support in operations maintenance of wind turbines.

The text mentions a video on YouTube that uses graph theory to explain the number of reasonable orders for watching Marvel's 23-movie Infinity Saga, discovered by the author while searching for WandaVision Easter eggs. The video is part of the Neo4j YouTube channel and provides an explanation of the basics of graph theory in a logical manner. The author encourages viewers to subscribe to the channel for more graph-related content, which is updated weekly with "tons of graph tech goods".

The presentation by Neerav Vyas, Head of Analytics at Realogy Holdings Corp, discusses how the real estate landscape is complex and fragmented, with multiple brands and legacy systems. He shares his personal experience of moving to New York City and how he realized that the journey of buying or selling a home is disconnected and traverses across different devices, channels, and domains. Vyas explains how using Neo4j has helped Realogy Holdings Corp match customers with the right agents for the job, drive lead conversion, and improve operational performance. He highlights four key areas where graph databases can drive value: data aggregation and curation, machine learning and AI, users, and leads use case. The presentation showcases the power of graph databases in transforming business operations, recruiting, and employee management.

Graph databases are being used by companies to solve complex problems, such as fraud detection in financial services. Traditional methods of fraud detection often fail to minimize losses due to false positives and negatives, but graph technology offers a new approach through advanced contextual link analysis, which can uncover patterns that are difficult to detect using traditional representations. A Fortune 500 financial services company used Neo4j to solve this problem by reducing the time it took to process fraud detection queries, providing real-time results with connected data and data visualization, and enabling analysts to make faster, more accurate decisions. This has resulted in a significant reduction in wait times for non-fraudulent customers and a potential saving of thousands of dollars per day. Graph technology is essential for detecting elaborate fraud rings that traditional technologies are not designed to handle.

The Neo4j community has been active with various releases, including a new version of the Neo4j Desktop and improvements to the Power BI connector. Alistair Jones launched a brand new version of the popular graph drawing tool, Arrows.app, with enhanced features such as easy node dragging and styling options. The community also explored the MET Art Collections using GraphAware Hume, and Antonin Smid continued his series on how to analyze the collections. Additionally, Tomaz Bratanic shared insights on using GraphSAGE embeddings for downstream classification models. The Developer Relations team highlighted Brandon Campbell's work as a featured community member, who presented at the NODES 2019 virtual conference. The team also announced new online training courses and thanked Brandon for his contributions to the Neo4j community.

Enel X, a smart energy leader, uses Neo4j to model its IoT devices and their relationships with buildings, customers, and sites. The company leverages Neo4j for fine-grained access control through its integration with ForgeRock, a product that provides access control information as a database.

The Neo4j Browser is a powerful tool for developers to write Cypher queries and interact with a Neo4j database, but it can sometimes cause frustration when query results take a long time to render. The issue may not be with the query itself, but rather with the rendering of the query results in the browser. The Neo4j Browser transforms multi-dimensional overlapping path data into a 2D visual, which can obscure the actual amount of data returned. To determine if the visualization is causing delays, developers can use techniques such as the `count()` test to understand how much data a query is returning and getting an idea of the query performance characteristics. Another option is to use Cypher-shell or Neo4j Bloom, which are alternative tools that allow for more efficient rendering of query results. Developers should also be mindful of their schema design and monitor their database usage to ensure optimal performance. Finally, it's essential to identify a graph use case and follow good data modeling techniques to avoid unnecessary complexity in the browser.

We have reorganized our Introduction to Neo4j course into multiple courses, covering various topics such as an overview of Neo4j, querying with Cypher, creating nodes and relationships, using indexes and query best practices, and importing data into the database. The courses are designed to be taken in a specific order and require a supported development environment. Each course is tailored to provide hands-on exercises and cover essential lessons for users who want to learn Neo4j without necessarily learning how to write Cypher statements or update the graph. By completing these courses, users can gain exclusive access to advanced training half-day courses upon passing the free certification exam, becoming a Neo4j Certified Professional.

Graph databases focus on data relationships, whereas SQL and NoSQL databases prioritize data aggregation and collection. Graph databases store data as nodes connected via relationships, allowing for efficient querying and navigation of complex data structures. These systems provide benefits such as simple and natural data modeling, flexibility for evolving data structures, simultaneous support for real-time updates and queries, and better query performance with connected data. In contrast to traditional rigid systems, graph databases offer a scalable, secure, and flexible platform for enterprises to deploy use cases, including real-time recommendation engines, fraud detection, knowledge graphs, and AI/machine learning applications.

The Neo4j community has released the first version of TWIN4j for 2021, featuring various video content including an interview with Emil Eifrem about the FinCEN Files investigation and projects by Antonin Smid, Dan Flavin, Sebastian Daschner, and Shyam Pratap Singh. Mark Heckler is highlighted as a featured community member, known for his work on Neo4j and Spring Boot. The community has also released new training courses, including an overview of Neo4j 4.0, and explored various topics such as analyzing The MET Art Collection, building a Christmas tree using Cypher, and decoding COVID-19 tweets using NLP and Neo4j. The community is actively engaging with its members through videos, blog posts, and social media, sharing knowledge and experiences on graph databases and related technologies.

In this update, we're discussing the latest GOTO Book Club episode featuring Jim Webber and Nicki Watt, who shed light on how to leverage graph databases to further understand data. The episode provides a comprehensive journey through graph technology, covering its definition, various use cases, and common pitfalls. To learn more about graph databases, add it to your quarantine reading list and catch all our videos by subscribing to the Neo4j YouTube channel.

We have designed a Neo4j data integration pipeline to streamline our projects, providing access and transparency to the entire process. The pipeline uses Snakemake rules to control each step of the build process, running checks on each dataset and automating the build process. It can create a working graph from raw data, while also handling datasets from various sources that require cleaning and QC before incorporation. The pipeline also includes features such as predefined database schema creation, testing new data, merging nodes, Neo4j import, remote server options, and setup instructions for use. Our goal is to provide a simple method for adding new data to a graph build, which could potentially be used collaboratively.

In a paradigm shift, organizations are moving away from traditional single-data models to more flexible data management approaches. Graph databases, which store and query data as nodes and relationships, offer a promising solution for managing complex, interconnected data. With the right approach, graph technology can help businesses unlock new insights, improve decision-making, and drive growth. To get started with graph technology, organizations should begin by identifying their specific use case or "graphy problem," building a proof of concept (POC) on a laptop, and then scaling up to a cloud pilot. By leveraging Neo4j, a popular graph database platform, businesses can unlock the full potential of connected data and drive business value through advanced analytics, recommendations, and new experiences. Ultimately, the key to success lies in understanding how graphs are used to solve real-world problems and driving meaningful conversations with stakeholders about the benefits and opportunities presented by this technology.

We love books. In that spirit, here are three groups of seven reading recommendations for 2021, covering topics such as fun and learning, graph databases and data science, and artificial intelligence. Seven books for fun and learning include Friend of a Friend by David Burkus, The Book of Why by Judea Pearl and Dana Mackenzie, Brains on Fire by Robbin Phillips, Greg Cordell, Geno Church and Spike Jones, Linked by Albert-László Barabási, World of Wonders by Aimee Nezhukumatathil, and Weapons of Math Destruction by Cathy O’Neil. Seven books for learning about graphs include Graph Databases For Dummies by Dr. Jim Webber and Rik Van Bruggen, Graph Data Science For Dummies by Amy Hodler and Mark Needham, Fullstack GraphQL Applications with GRANDstack by William Lyon, Graph-Powered Machine Learning by Alessandro Negro, Graph Algorithms: Practical Examples in Apache Spark & Neo4j by Mark Needham & Amy E. Hodler, and AI on Trial by Mark Deem and Peter Warren. Seven books about AI include You Look Like a Thing and I Love You by Janelle Shane, The Algorithm Design Manual by Steven Skiena, Reprogramming the American Dream by Kevin Scott and Greg Shaw, Deep Medicine by Eric Topol, Hello World: Being Human in the Age of Algorithms by Hannah Fry, and The AI Book by Susanne Chishti, Ivana Bartoletti, Anne Leslie and Shân M. Millie.

The Neo4j community has been actively sharing their knowledge and experiences through various podcasts in 2020. The top episodes of the year include interviews with influential figures such as Alexander Jarasch and Martin Preusse, who co-founded CovidGraph to fight the COVID-19 pandemic using graph technology. Other notable episodes feature discussions on common use cases of graph database technology, the basics of graph data visualization, and insights into the future of graph data technology from Neo4j CEO Emil Eifrem. The Graphistania podcast has also covered topics such as the defining characteristics of a graph database, common pitfalls to avoid, and the growing field of graph data science. These podcasts offer valuable information and insights for both technical and non-technical listeners interested in learning more about graph databases and their applications.