September 2021 Summaries

The text highlights key themes in the tech industry, including explainable AI, low-code/no-code movement, space exploration, and graph technology. Explainable AI has emerged as a crucial area of focus, with AI providing wrong answers and a need to make predictions easier to trace and explain. Graph technology is seen as a solution, enabling organizations to understand complex data relationships better. The low-code/no-code movement aims to democratize coding, making it accessible to more people, including "citizen developers" who can help build applications quickly. Space exploration is also on the radar, with a focus on safety, security, and sustainability in space. Companies like Neo4j are leading the graph technology revolution, with its database market expected to reach $100 billion. The industry is expected to continue growing, with graph technology playing a crucial role in solving complex problems and understanding data in new ways.

Neo4j and the Cloud digital Connections event provided a platform for thought leaders and experts to share their experiences with graph databases. Boston Scientific's Eric Wespi discussed how his team uses Neo4j in the manufacturing organization, focusing on quality investigations, yield improvement, and supply chain optimization. He highlighted the benefits of using a graph database, including speed, standardization, and finding hidden insights. The company recently joined AuraDB Enterprise, a managed cloud solution, to focus on data model analysis and movement of data in and out, rather than administrative tasks. Eric shared his positive experience with AuraDB, citing its scalability, support, and ease of use. He emphasized the importance of engaging practitioners and empowering them to explore new ideas and projects, creating an engine for innovation that drives creativity and new use cases.

Hello, everyone! This week we're featuring Anthony Chiboucas, a very active Neo4j Community member. He's been working on a graph-enabled middleware solution to connect artists, places, events, and genres throughout all of history for museums and arts exhibitions. GraphAware and Breadcrumbs collaborated on a project to analyze a graph with billions of crypto transactions and visualize the data. The folks at Osinto – who provide subscriptions to their knowledge graphs – also have a product where you can visualize data and relationships in 3D using Cloud Empire. Jennifer Reif loaded a Neo4j database with Java library data, and she demonstrates how to analyze the differences between Java versions. Meet Suvariya wrote a beginner’s tutorial aimed at data scientists exploring the graph, and he also shows us how to use Gephi to visualize the data. Neo4j engineers released version 2.2.0 of the GraphQL Library. And Adamantios – Marios Berzovitis released GraphKer for performing cybersecurity checks on your graphs. Finally, we leave you with links to the most popular video series produced by the Developer Relations team.

This post from Neo4j's developer relations engineer discusses the various temporal functions available in APOC, a library for extended Cypher functionality. The author reviews each function, comparing it to its equivalent in Cypher and explaining when to use each. Key points include: APOC's `apoc.date.*` functions can be replaced by Cypher's built-in temporal functions in some cases, but are still valuable for dealing with epoch time and strings; the `apoc.date.parse()` function is simpler than its equivalent in Cypher for converting a string to a specific time unit; the `apoc.date.parseAsZonedDateTime()` function returns an ISO8601 temporal value, whereas Cypher's temporal instants do not account for timezone offsets; APOC's `apoc.temporal.*` functions are used to format temporal values into strings in different formats, which is not possible with Cypher. The author concludes that while some APOC functions can be replaced by Cypher's built-in functionality, others remain valuable and should be used when necessary.

The latest release of Neo4j's Graph Data Science (GDS) focuses on making graph data science accessible, easy, and foolproof. It introduces machine learning pipelines for graph native link prediction, allowing users to define the steps they want to take to build their predictive model without worrying about data manipulation and math. This feature enables users to create a pipeline, define node and link features, split the data, and select the best-performing model, while Neo4j takes care of assembling those steps in the right order. The release also includes progress logging and system monitoring to make model building easier and less error-prone, as well as three new features requested by users: Cypher on the GDS Graph, string support for graph export, and graph partitioning.

Neo4j has released version 4.1 of its AuraDB, which includes a new Kafka Connect Source connector that enables bi-directional communication between Neo4j and Apache Kafka. This release allows customers to produce the results of Cypher queries to any Kafka topic, making it simple to quickly define connectors that move large data sets into and out of Kafka. The new functionality uses Kafka Connect, a tool for scalably and reliably streaming data between Apache Kafka and other data systems. With this release, Neo4j AuraDB customers can now have full bi-directional support to and from Kafka, unlocking powerful use cases such as flexible fraud detection. The connector works by defining a polling query and interval, handling fetching all data that has changed since the last poll, and producing all of the resulting information to the topic of your choice. This release also deprecates the Neo4j Database plugin in favor of the Kafka Connect approach, which offers several advantages such as easier installation, configuration, and monitoring. The best place to find out more about this release is via the official documentation, or by reaching out to a Neo4j representative.

The new version of Neo4j GraphAcademy is now available, offering a free, self-paced online training platform with five beginner courses. The courses aim to provide comprehensive content completely free of charge, but some users found the site overwhelming due to its extensive course offerings and lack of clear guidance on where to start. To address this, the new version features improved navigation, shorter course lengths, interactive code challenges, a recommended learning path towards certification, and the ability to share progress and earn digital rewards, including limited edition t-shirts. The platform is designed to work seamlessly with Neo4j Aura, the company's Database-as-a-Service platform, and users can provide feedback on the new courses through built-in widgets.

The Neo4j community has been actively sharing knowledge and resources. The developer relations team recorded hands-on training sessions, which were made available to the public. The Kineviz Community created a tutorial on using GraphXR for visualization of data in a Neo4j Sandbox without requiring any installations. Sixing Huang shared an article about exploring the Comprehensive Antibiotic Resistance Database (CARD) using Neo4j. Lju Lazaravic published a tutorial comparing relational and graph data modeling, including how to access movie data using Cypher versus SQL. Charlotte Skardon wrote about reusing Neo4j driver instances in Azure functions with Dependency Injection. Jennifer Reif continued her blog series on migrating from Spring Data Neo4j 5 to 6, providing guidance on using the migration repository for upgrades. David Allen shared a quick guide on loading JSON data into Neo4j using the APOC library. The Developer Relations team also recorded several video sessions covering various topics such as AuraDB Free and Knowledge Graphs.

If you're considering implementing a knowledge graph, ask yourself if your data is highly related, how often you join data and what databases you already have. You should also think about the time, money and resources required to maintain it, as well as whether your staff has the necessary expertise and if you already have graph or graph-like data. Additionally, consider downstream systems that will use graph data and how you plan to keep the data up-to-date. Understanding these factors can help determine if a knowledge graph is right for your organization.

The text discusses exploring the newly-released Java 17 version using a graph database. The author, a Developer Relations Engineer at Neo4j, uses a JDK data set that details historical library changes of Java versions to create a graph model and perform queries to analyze the changes between Java 16 and Java 17. The graph model represents Java versions as entities with diffs, forming a tree-like structure when visualized. The author creates Cypher statements to import the data into Neo4j and runs queries to find changes in specific packages, such as java.time, and to identify new features and functionality added in Java 17. The analysis provides insights into the changes made in Java 17 and offers opportunities for further exploration, including modeling alternative approaches and building a Spring Data Neo4j application.

The Neo4j GraphConnect Awards, also known as the "Graphies," are an annual celebration of the best in the connected data space. The awards recognize individuals and projects actively contributing to making the world a better place with graph technology. Past winners include notable companies like Meredith, Lockheed Martin Space, and Ebay, as well as community partners such as EY and GraphAware. The awards aim to showcase innovative work using Neo4j, and nominations are now open until September 17th. To be considered, applicants must provide a brief overview of their project, explain the problem they're solving, why they chose Neo4j, and share results and supporting materials. The new Sr. Customer Advocacy Manager, Juan Ortiz, joins the team this year, bringing experience from previous roles in customer advocacy and executive programs.

Data modeling has become outdated, with many recent developments and research studies failing to produce new data modeling books. The main purpose of a data model is to ensure consistent storage and retrieval of data from databases according to business rules, while also facilitating efficient communication among different individuals and teams through visualizations. Effective visualization fosters high-level cognitive abilities such as classification, unification, knowledge completion, and reasoning. However, traditional relational databases with rigid schemas are not well-suited for rapid change, whereas graph databases can efficiently store and query connections between entities, making them ideal for complex, interconnected data. Graph models offer a "whiteboard-friendly" approach to modeling, where the relationships between data are as important as the data itself, allowing for intuitive communication and learning. The concept of whiteboard-friendly modeling involves using three independent schemas: conceptual, logical, and physical, with the latter often being just the database schema.

This week marks the return of many community members from their time off, and a new team member joins Neo4j. The training week for September is coming up, covering topics such as Neo4j, AuraDB Free, Bloom graph visualization, GraphQL integration, Knowledge graphs using NLP and GDS, and more. AppsFlyer shares its approach to securely managing microservices with policy-based access control (PBAC) using Neo4j. Suadeo, a mobile recommendation engine app based on Neo4j, is released by community member mbazos. Ghislain Atemezing from Mondeca shares slides from Semantics Conf on how they publish and browse knowledge graphs using Neo4j. Two favorite contributors explore their favorite graph algorithms in a joint video. Bryant Avey explains how to surface patterns detected in a Neo4j graph automatically in Power BI. The GDS library 1.7.0 is released with new maximum k-cut algorithm, ML pipelines, Cypher on the in-memory graph, and progress/monitoring improvements. Various quick hits and releases are shared, including updates from GraphAware, LinuxHint, and others.

Knowledge graphs have emerged as a crucial technology in recent years, leveraging machine learning and artificial intelligence to make data more reliable, robust, trustworthy, and explainable. They organize data with context and relationships, driving intelligence into data via an "organizing principle" that adds meaning to the data. Knowledge graphs are used across various use cases, including data governance and compliance, risk management, augmented MDM, X-360, AML, root cause analysis, and many others, providing data assurance and radical visibility. They also play a key role in contextual AI, making AI more trustworthy, accurate, and allowing for better reasoning. Furthermore, knowledge graphs are transforming businesses by bridging data silos, building a data fabric, accelerating machine learning and AI adoption, and providing a blueprint for digital twins. The book "Knowledge Graphs: Data in Context for Responsive Businesses" offers a practical guide for business leaders, data scientists, and developers to harness the power of knowledge graphs.

Neo4j, a leading graph database, has released updates to improve performance, scalability, security, and developer experience. The latest release, Neo4j 4.3, offers significant enhancements, including improved performance, unlimited graph-native scale, fine-grained security, and cloud-scale capabilities. A new visualization tool called Neo4j Bloom streamlines conversations and projects across teams, empowering non-technical participants to share in innovative work. Additionally, a webinar series covered the latest releases, including a session on Neo4j DevTools, which provides flexible developer tools for building graph applications with ease. The updates aim to support developers in building intelligent applications with rapid, agile development while integrating languages, libraries, and APIs seamlessly.

Twin4j` is a community-driven project that aims to provide a unified interface for Neo4j, a graph database. The project has several exciting updates and content this week, including live re-runs of the `NODES 2021` workshops, knowledge graphs, data lineage research, and a learning experiment to create a social network for bunnies. The community is also welcoming new member Adnan Siddiqi, an independent software consultant who recently explored Neo4j and created content on visualizing Python modules and dependencies with the graph database. Additionally, Twin4j is hosting a Neo4j Training Series in September, covering topics such as AuraDB, Bloom, GraphQL, and Knowledge Graphs. The community is also sharing fun projects like `Planning a Trip Through the State of New York` using AuraDB Free Tier, enabling data lineage, LinkedImm graph database for immunological data, and Bunnybook social network built with Neo4j, FastAPI, React/RxJs, PostgreSQL, and Redis.

Neo4j Bloom is a graph communication and data visualization tool that empowers developers, data scientists, and analysts to build intelligent applications and machine learning workflows. It provides a breakthrough visual interface for users to share innovative work, streamlining conversations and projects. The latest release of Neo4j Bloom 1.8 introduces several key enhancements, including improved support for date/time properties, export screenshot capabilities, multi-user editing of perspectives, and Single Sign-On (SSO) login options. These updates aim to make it easier for users to explore their data, identify patterns, and share findings with others.