November 2020 Summaries

The Neo4j BI Connector is a product that delivers direct access to Neo4j graph data from business intelligence (BI) tools, such as Tableau and Looker. It's the first enterprise-ready, supported product to do so without requiring coding or custom scripting. The connector uses a regular instance of the Neo4j official Java Driver / Bolt client for Neo4j and supports different connection schemes, including identity separation, security, query throttling, and more. The BI Connector also introduces Neo4j Bloom, a graph-native visualization tool that allows users to visually investigate and explore their graph data from different business perspectives. While the connector has some advantages, such as being accessible to business analysts and based on SQL technology, it also has disadvantages, including limitations in visualizing "graph native" features like Path Finding. The connector can be used with batch or live pull operations, depending on the size of the database and query frequency. Currently, the connector is JDBC-only, but an ODBC version is planned for the future. Overall, the Neo4j BI Connector bridges the gap between familiar BI tools and graph data platforms like Neo4j.

The Developer Relations team at Neo4j has released several new content pieces, including a video on using the new JSON Bolt serialisation format with Cloudflare Workers, and blog posts on how to test Spring Data Neo4j components and export Spotify playlists into Neo4j. A featured community member this week is Will Reynolds, who discusses his work in revolutionizing the construction industry with graph data. Additionally, there are updates on the Perl driver, APOC procedures, and RFM analysis using e-commerce data. The team also shares a tweet of the week from Kristijan, expressing enthusiasm for Neo4j and its Cypher language.

Seyed Hossein Ahmadinejad, an expert in identity and access management, believes that as more people recognize the significance of relationships between identities and connections among them, they will migrate from traditional databases to graph data platforms like Neo4j. He has been working on security, privacy, and identity and access management since 2008 and has experience deploying over a billion identities for clients globally. Ahmadinejad's company, Nulli, uses innovative ideas and solutions to address the challenges of identity and access management, which he attributes to the need to model complex relationships between entities in data storage. He notes that traditional relational databases are not scalable for this purpose, but graph databases like Neo4j can effectively navigate these relationships, making it a suitable choice for his company's projects. Ahmadinejad also uses Neo4j for data analytics and has successfully applied graph algorithms to simplify the management of complex entitlements and roles in identity and access management systems. He believes that as more people recognize the importance of relationships between identities, they will increasingly adopt graph databases like Neo4j for their security and access needs.

We have updated a course in our catalog of free online courses – Using a Machine Learning Workflow for Link Prediction, which is intended for experienced Cypher and Python developers and data scientists who want to learn how to apply graph algorithms from the Neo4j Graph Data Science Library using a machine learning workflow. The course uses Neo4j 4.0 in a Neo4j Desktop environment, with most of the hands-on coding done in Jupyter notebooks run on local systems. The course covers topics such as setting up the development environment, exploratory data analysis, building a recommendation engine for authors, and learning how to build a machine learning classifier to predict co-authorships in the citation graph. The course includes four hours of hands-on exercises and is an update to the previous course Data Science with Neo4j 3.5.

The Neo4j BI Connector is a tool that enables direct access to Neo4j graph data from business intelligence (BI) tools, such as Tableau and Looker. It allows users to avoid coding and custom scripting, while also providing connected data results. The connector can be used for various use cases, including money laundering detection, business impact planning, network and IT operations, product recommendations, and building a promotional discounting program. By using the Neo4j BI Connector, businesses can gain insights into their complex data structures and make informed decisions.

The Developer Relations team at Neo4j has been busy adding usage examples to the APOC documentation, with over 340 functions and procedures covered so far. The team also released Neo4j 4.2 and Hume 2.6.0, which include enhanced Cypher planner, improved OLTP optimisation runtime, as well as enhanced security features. In addition, there are several blog posts available on topics such as the Graph Data Science Library and data auditing with Spring Data Neo4j. The team also featured community member Julia Neagu, who presented a session on building an anti-money laundering solution using Tamr and Neo4j. A new video is available on building a graph visualization using PIXI.js, and the team is exploring the k-nearest neighbors algorithm in the Graph Data Science library.

In a recent video, David Allen demonstrated the use of Neo4j BI Connector to feed live graph data into Tableau for product recommendations. The connector enables users to visualize and analyze customer clusters in real-time using graph technology. This integration showcases how Neo4j can be used with other tools like Tableau to provide more insights and applications.

Graph data science is a rapidly growing field that uses graph-based analysis to explore deeper meaning in existing data, improve forecasts, and make better predictions. The increasing connectedness of data, breakthroughs in scaling graph technology, and the integration with machine learning (ML) and artificial intelligence (AI) solutions have contributed to its accelerating adoption in business. Graphs are a mathematical abstraction of complex systems, representing entities as nodes and relationships between them as lines. They can be used to answer tough questions, such as how things move through a network, what are the most influential points, and what patterns are significant. The field of graph data science encompasses three main areas: graph statistics, graph analytics, and graph-enhanced ML and AI. These areas aim to provide insights into relationships and structures in data to power predictions. With the rise of graph technology, organizations are increasingly using graph data science to gain knowledge from their data, and research has seen a significant increase in the use of graph technology in AI research over the past decade.

The new Spring Data Neo4j 6.0 is a major release built on the foundation of SDN/RX, which was released earlier this year. It addresses issues with Neo4j-OGM by providing full support for reactive programming, mapping layer compatibility, and improved driver support. The new release introduces a record-oriented mapping approach, supports findByExample methods, and features a Cypher-DSL for custom queries. SDN 6 is designed to be thin client over the driver, allowing users to manage transactions and provide hooks for their own mapping. It also includes configuration support for the Neo4j Java Driver in Spring Boot, better health endpoint information, and potential Micrometer metrics support. The release provides a stable framework for the next few years and encourages feedback from the community.

Neo4j 4.0 marked a significant milestone in graph technology, emerging the graph database into the mainstream of enterprise technology. Neo4j 4.2 builds upon this success by completing critical capabilities in performance, operability, and security. The release brings substantial improvements across the stack, including enhanced Cypher planner, parser efficiency gains, and increased sharding accessibility. Additionally, operability enhancements focus on multi-database instances, multi-tenancy scenarios, and better portability of permissions. Security features are also strengthened with extensions to role-based access control and internal/user-defined procedures and functions. With the release being the most stable and battle-tested yet, Neo4j 4.2 is now available for download, offering a quick getting started experience.

The Neo4j BI Connector enables direct access to Neo4j graph data from business intelligence (BI) tools, providing a seamless and real-time experience. It connects to Neo4j databases in real time, allows for querying using SQL, and makes connected data insights accessible. The product is enterprise-ready, supported, and available at no extra charge for Neo4j Enterprise Edition customers. It bridges the gap between BI tooling and graph databases, making it easy to combine graph insights with existing decision processes and drive business intelligence. The connector also facilitates low-code/no-code integration, allowing non-graph users to access graph data without custom code, and enables sharing of data between organizational units in a simple way, putting Neo4j on equal footing with other databases or data warehouses.

The Neo4j community has been actively engaged this week, with various developers and experts sharing their knowledge and experiences. Adam Cowley showcased the React Hooks For Neo4j library, allowing users to build graph-based applications more easily. Kristof Neys compared NetworkX and Neo4j's Graph Data Science library in a "drag race" of sorts, highlighting their strengths and weaknesses. Hantsy started writing blog posts on using Neo4j with Spring Boot 2.4, while Maya Natarajan and David Allen announced the launch of the Neo4j Connector for Apache Spark. Meanwhile, Luanne Misquitta, Vice President Of Engineering at GraphAware, was featured as a community member, having been involved in early work on Neo4j OGM and Spring Data Neo4j. Additionally, there were updates on Neo4j AuraDB's launch on Google Cloud Platform Marketplace and the Graphistania podcast discussing various topics related to graph databases.

The Innovation Lab is a concept developed to bring together data science and design thinking to quickly prototype and validate new use cases. A diverse group of people with different backgrounds, ages, and qualifications are brought together to create a remix of their mental paradigms. The end result of a successful Innovation Lab is the validation of a use case and finding something of real value. Graph technology enables the connection of seemingly unrelated things, which is how innovation works. To get started with graphs, one needs to start doing it by trying out tools like Neo4j and actually working on projects. Removing friction and making it easy for people to try new things is part of the goal of the Innovation Lab.

To create successful graph applications, it's essential to use the right tool for the job, leverage connections and speed, and start with obvious use cases like social networks or knowledge graphs. Good graph-based systems begin with modeling, starting small and scaling later, and focusing on value rather than just technical aspects. Graphs are valuable when used to find patterns and make predictions about future dynamics of the network, and exploring hidden insights can reveal new information. Finally, connecting with the graph community is crucial for learning from others and realizing the value of graph technology more quickly.

Neo4j Connector for Apache Spark is an integration tool that enables bi-directional movement of data between Neo4j graph database and Apache Spark, a powerful ETL tool. This connector allows users to leverage the strength of both platforms, with Neo4j providing advanced graph functionality and Spark offering data orchestration capabilities. The connector uses Neo4j's official Bolt driver for Java, making it easy to connect and work with nodes, relationships, and queries in both directions. It is ideal for ETL work with Neo4j, enabling users to assemble and transform data, reshape tables to graphs, enrich data, and further process insights derived from graph analytics. The connector is also suitable for data scientists and developers who need to manipulate data between Spark and other sources, such as Oracle or Snowflake.

As a veteran with over five years of cumulative active duty time, Cory Waddingham has gained valuable experience as an Operations Specialist in the U.S. Navy and later as an IT Specialist. His service was motivated by family tradition and a desire to see the world. Running tactical networking systems like NTDS on his first ship set him up for a successful career in tech, where he now serves as Lead Support Engineer for Neo4j AuraDB. He attributes his leadership skills to his experience working with fellow Navy personnel towards a common goal, sacrificing personal comfort for the greater good. Cory believes that this spirit can be applied in corporate environments, fostering strong and successful companies.

Rik Van Bruggen explores the hundreds of Zoom calls he's been in since Lockdown #1, while Florent Biville builds a Realworld.io example using Neo4j and Go. Kristof Neys uses Graph Neural Networks to approximate network centralities in Neo4j, and David Allen explains how queries work in Neo4j. Liz Maida, founder of Uplevel Security, is featured as this week's community member, sharing her experience with Neo4j and graph data modeling. The Neo4j Commander tool has released version 2.4 with syntax highlighting, and a new Udemy course on building an IMDB clone with Python and Neo4j was launched. Additionally, the first NODES 2020 Extended session featured talks on Contentful and Enterprise Application Integration with graphs.

The discussion focuses on Pathfinding algorithms, specifically the Shortest Path algorithm, which calculates the shortest path between two nodes in a graph. The algorithm can be applied to weighted or unweighted paths, with the former taking into account costs and the latter being determined by the number of relationships between nodes. The concept is relevant to various applications, including biotechnology and pharmaceutical industries, as well as social media platforms like LinkedIn and mapping services like Google Maps.

Financial fraud is becoming increasingly sophisticated, and to combat it, you need a better understanding of your data. Graph data science can help uncover hidden patterns and connections in your existing data, making it a valuable tool for detecting and preventing costly fraud schemes. By tapping into these patterns and connections, you can gain a more accurate understanding of your data and make informed decisions about how to protect against financial fraud.