February 2022 Summaries

Neo4j is a graph database that fits into the modern cloud ecosystem as part of full stack development, particularly in operational workloads. It can be integrated with other technologies through an API layer to handle tasks such as data modeling and loading. The Neo4j Graph Platform offers various tools for developers, including Arrows for graph diagramming and the APOC procedure library for working with JSON data. GitHub Actions can be used to automate tasks such as loading data into Neo4j AuraDB instances. The Neo4j GraphQL Library provides a no-JS solution for building JavaScript GraphQL APIs, which can be used to expose application data to frontend applications. A React app built using Next.js and the Apollo Client library can be used to visualize graph data in real-time, with GitHub Actions continuously updating the data in the background. Deployment of the application is done through Vercel, a platform optimized for developers. The application was built on Neo4j Livestream, where live coding and Q&A sessions are held. Resources such as Neo4j AuraDB free tier, the Neo4j GraphQL Library, GraphAcademy, and job openings can be found online.

This week GraphAware launched the new no-code Neo4j Data Importer, which allows users to easily import data into their AuraDB database or self-hosted Neo4j instance. The tool uses a simple 3-step process: loading data from flat files, defining a graph model and mapping data to it, and importing the data into the chosen database. To use the Data Importer on a self-hosted Neo4j instance, users can visit a specific URL for non-secure or secure instances. GraphAware also provides additional resources, including videos and subscriptions to their YouTube channel, which is updated weekly with graph technology content.

The COVID Craft Initiative, led by HealthECCO and the German Center for Diabetes Research, is harnessing graph technology to aid in research against COVID-19. The team uses Neo4j to connect data from various sources, including patterns, publications, and molecular data, to create a comprehensive knowledge graph. This initiative has won a 2021 Graphie Award and has been recognized as a groundbreaking medical research project. The team's backgrounds in bioinformatics and computer science have empowered their work, with Alexander Jarasch bringing his expertise in managing data and Martin Preusse leveraging his experience with Neo4j to drive the project forward. They highlight the ease of integrating data, visualizing results, and facilitating communication among researchers as key benefits of using graph technology. The team projects that graph technology will play a significant role in future medical research, particularly in knowledge graph development and graph-native learning.

The Neo4j company has announced the beta release of its new Data Importer, a no-code user interface for importing flat file data into a graph database. The tool is designed to quickly load 2-dimensional data files but does not replace existing ETL tools or handle production loads. This new feature was added in response to user feedback on the Neo4j Browser, which includes trackpad zoom gestures and a fit-to-screen feature that enhances graph visualization. Additionally, other releases were announced, including updates to Bloom and APOC, as well as a GraphQL library release, a data science article featuring Dr. Jim Webber, and a tutorial on creating a graph for a Kaggle competition. The company is also encouraging users to participate in its Neo4j Developer Survey and calling for submissions for the GraphConnect 2022 conference.

The Neo4j DevTools team has released their first updates of the year, including a beta launch of the Neo4j Data Importer and new productivity features for Neo4j Browser. The Data Importer is a no-code UI for loading flat file data into Neo4j databases, supporting up to 1 million rows per input file, graph modeling, and mapping data to structure and properties. It's available now in AuraDB and can be accessed directly from the database settings. Additionally, Neo4j Browser has received updates with new zoom controls and interactions, including trackpad zoom gestures and a fit-to-screen feature. The team has also made under-the-hood upgrades to the d3 visualization library and introduced parameter assistance to help users construct more efficient Cypher queries.

This week marks the release of Neo4j GraphQL Library version 3.0.0, a major milestone in integrating GraphQL with Neo4j, enabling developers to leverage graph data throughout their stack. The community has been actively contributing to the library, making it easier for developers to build GraphQL APIs backed by Neo4j. The latest release focuses on improving working with relationships in GraphQL and provides best practices for using Neo4j drivers. Various resources are available for learning about Neo4j and GraphQL, including a top-rated course on Udemy taught by Louis Sayers, a blog post announcing the new features, and a limited edition t-shirt giveaway for completing a course on Graph Academy.

Neo4j Bloom 2.1 has been released with a new feature that allows filtering and rule-based styling of nodes and relationships based on times and dates, including handling time zones. This feature is particularly useful for analyzing network traffic data from different locations with varying time offsets. With this feature, users can now filter and style their data in a more intuitive way, taking into account the local time zone where the data was collected. The feature works seamlessly with both `time` and `datetime` values, allowing users to translate time zones and adjust the display of time values accordingly. This enables users to analyze their data across different time zones in a more meaningful way, without having to normalize all times to a common offset like UTC.

Neo4j is excited to announce the release of the Neo4j GraphQL Library 3.0.0! Initially driven by the requirement to perform some major dependency upgrades, we have taken the opportunity to work through a backlog of breaking changes to improve the experience with the library, and to perform some much needed cleanup in key areas. The new version introduces several new features, including support for undirected relationships traversals, relationship cardinality enforcement, and new filters for querying data. Additionally, query limit control has been added, allowing users to set a default and maximum number of entries returned for a particular type. The library also includes improvements to bug triage and deprecation handling, as well as several minor bug fixes. Users can now expect more confidence in the integrity of their relationships and enjoy new ways of querying them entirely.

The author reminisces about the winter of 2020 when Neo4j was growing rapidly, and they were organizing "GraphTour" events across Europe. However, with the COVID-19 pandemic hitting in early March 2020, things took a dramatic turn, and the company had to adapt. As a response to the lockdown, the author started a daily call with colleagues from different countries and time zones, which became known as the "Breakfast Club." The call was informal, without an agenda or leader, but provided a sense of community and connection among participants. Over time, it led to the formation of real friendships and reinforced the company's values of relationships and support. The author reflects on how this experience not only helped them cope with the pandemic but also strengthened their culture and sense of belonging within the organization.

The Neo4j driver best practices emphasize the importance of using secure connections, such as `neo4j+s://`, which provides certificate validation and ensures a secure connection. Using very recent versions of Neo4j drivers is also crucial, particularly for Neo4j Aura environments. It's essential to verify connectivity before issuing queries and create one driver instance once and hold onto it. Explicit transaction functions are recommended over auto-commit transactions, as they provide better control over commit behavior and allow for more efficient query execution. Query parameters should be used whenever possible to pass variables into Cypher queries, avoiding string concatenation. Processing database results within the transaction function is also important to avoid issues with result objects being available outside their scope. Bookmarking and causal consistency are key concepts in Neo4j drivers, allowing for reads of own writes and ensuring consistent reads across cluster members. By following these best practices, graph code can be made safer, more secure, more performant, and more portable between different graph environments.

Neo4j's Graph Data Science framework is an approach to resolve data into unique and valuable entity profiles, addressing the challenge of consolidating and disentangling multiple data streams to glean actionable insights. Entity resolution is crucial for understanding a business, particularly in resolving questions about people, places, organizations, and actions, leveraging traditional identifiers and behavior. The framework enables businesses to ingest massive data streams, resolve them into unique targeted profiles, and create descriptive audience segments leading to personalized experiences. By using Neo4j's Graph Data Science Library, companies can apply graph algorithms such as Node Similarity and Weakly Connected Components, and utilize supervised machine learning techniques like Link Prediction. This approach has been successfully implemented by Meredith Corporation, a media conglomerate with 180 million users monthly, resulting in a 612% increase in visits per profile. The framework provides fast and valuable results for various data challenges, including finding fraud, audience targeting, and ensuring correct customer data.

The text discusses a project to represent the Wordle word game as a graph in Neo4j, a graph database. The goal is to solve the puzzle using Cypher queries. Two models are presented: Model 1 represents characters at positions with dedicated nodes labeled CharAtPos, and Model 2 represents just the 26 characters and puts the position onto the relationship either as a property or as the rel-type. The text also explores follow frequencies, which can help in finding missing letters, and implements a generic solver that takes an structured input and includes or excludes letters based on their positions. A script is provided to play Wordle in the terminal by sending Cypher queries to a wordle database.

The Neo4j community is actively engaging in various livestreams and projects, showcasing their work in graph databases. The community has contributed several episodes of "Learn with Jason" where he teaches viewers about using a graph database, while also featuring regular streams such as "Going Meta", "Neo4j + PHP", and "Discover Neo4j AuraDB Free". Additionally, the team behind GraphAware created an interesting knowledge graph by analyzing their music tastes, which can be explored to discover new possibilities. Other notable projects include building a medical subject headings ontology in Neo4j and using Testcontainers to showcase how to store data in a graph database. The community is also celebrating its featured member, Cathereine Luse, who contributes to open-source software development with an ethos of contributing to a better world. Furthermore, the company behind Neo4j has made significant progress over the years, and their CEO Emil Eifrem discussed major milestones and advancements in the graph database field during a recent interview.

Spring Boot 2.4+ provides tailored-made Neo4j connectivity through its broad range of configurable properties, making it easier for developers to integrate the Neo4j Java Driver into their applications without writing extensive configuration code. The Neo4j team worked closely with VMWare engineers to ensure seamless integration. Developers can either add the driver dependency manually or use start.spring.io to access the driver and Spring Data Neo4j. With Spring Boot's dependency management, developers always have the latest, compatible version of the Neo4j Java Driver. In most cases, no code is required to configure the Neo4j connection, and essential properties such as URL, username, and password can be provided through simple configuration. However, in rare cases where custom configuration is necessary, developers can use the `ConfigBuilderCustomizer` or inject `Neo4jProperties` into a `@Bean` annotated method to tailor their configuration. The Neo4j team encourages developers to try out Neo4j AuraDB, their cloud offering that supports a broad range of ecosystems, including Java and Spring Boot.

PwC Canada has been honored with a Graphie Award for being a Pioneer in AML, Fraud, and Cybersecurity. Sam Samod, Partner at PwC, discussed his work using Neo4j graph technology to detect synthetic IDs and financial crimes. He revealed that they use Aura to prepare data and Bloom for visualizations, allowing them to correlate data from multiple sources effectively and quickly. Sam highlighted the benefits of graph technology in detecting bad people with high accuracy, cutting down investigation time by 80-90%, and providing a factual, unbiased approach to cybersecurity and fraud detection. He also emphasized the importance of leveraging consortium models and open banking initiatives to tackle financial crime prevention challenges. Overall, Sam believes that graph technology will play a significant role in shaping the future of financial crime prevention and cybersecurity.

Graph analytics are becoming increasingly important for enterprises to advance their AI enablement initiatives, but operationalization and post-production maintenance pose significant challenges. Enterprises often lack the required skill sets and tools in their current architectures, leading them to seek high-level architecture designs to get started with a minimal setup that can be scaled as needed. The Graph Analytics Platform architecture has been designed using Neo4j and GCP, which can also be deployed on other major cloud vendors or on-prem infrastructures. The Minimal and Scaled architectures are two design options for enterprises, with the former focusing on enriching data and generating insights and the latter enabling operational applications such as real-time fraud detection and border control. A key aspect of any modern Data Analytics Platform is the data life cycle, which drives the actions performed by different parts of the platform. The architecture diagrams illustrate the functions served by each section of the Graph Analytics Platform, providing a comprehensive solution for enterprises to advance their AI enablement initiatives.

The author of this article is a Developer Relations Engineer at Neo4j, who created a Java-based AWS Lambda function that connects to a Neo4j graph database and executes a query. The function uses the Neo4j Java driver to connect to a free cloud instance of Neo4j AuraDB, which contains a movie dataset. The author followed the instructions from an existing Neo4j Java example application and adapted it for use with AWS Lambda. They used JSON files for input and output, copied/pasted code where possible, and made only minor customizations. The function successfully executed a query that retrieves a person's name from the database and returned the result in a formatted JSON file. The author provides resources for further learning, including a GitHub project, an AWS Lambda example app, and Neo4j documentation.

The Knowledge Graph concept, introduced by Google in 2012, has become a powerful tool for linking data points to gain better insights and develop new applications. The Clinical Knowledge Graph with proteomics data presented in a Nature article is an example of its potential. To build your own Knowledge Graph, you can use the tutorial "Build a Knowledge Graph using NLP and Ontologies". GraphConnect 2022 has been announced, and researchers are encouraged to submit their work on graph-related topics. The Neo4j community member Dana Canzano is recognized for her patience and humility in helping others, and her contributions have made a significant impact on many people's lives. Meanwhile, various tutorials and courses, such as "Journeys in Java, Level 1: Building an Empire of Microservices", "Building Neo4j Applications with Python", and "A Comprehensive Guide on Neo4j" are available to help developers learn about graph databases like Neo4j. Additionally, the Mann labs' Clinical Knowledge Graph has been developed to analyze clinical proteomics data, while Michael Hunger and Alexander Erdl showcased how to analyze NFT trades using Neo4j AuraDB in a recent episode of Discover AuraDB.

Google launched the Knowledge Graph in 2012, connecting data points to gain insights and develop applications. The Clinical Knowledge Graph with proteomics data presented in a Nature article is an example of this concept. Building one's own knowledge graph can be done using NLP and ontologies, as demonstrated by Google. Dana Canzano is highlighted for her patience and humility in the Neo4j community, helping countless people. Jennifer Reif created two Spring Boot applications that communicated over HTTP to demonstrate building a microservices empire. A new course on building Neo4j applications with Python has been released, while Mrinal Singh Walia wrote a comprehensive guide on Neo4j for beginners. The Mann labs present a clinical knowledge graph with 20 million nodes and 220 million relationships, accelerating analysis and interpretation of proteomics workflows. Michael Hunger and Alexander Erdl analyzed NFT trades using Neo4j AuraDB in the latest episode of Discover AuraDB.

GraphAware showcased how they built a music knowledge graph using Neo4j as the database for workflow data processing. They collected almost 400 song recommendations from team members and created a database of their musical tastes. The project highlights the potential applications of graph technology in exploring individual interests and preferences.

The author of the text, a Head of Product Innovation & Developer Strategy at Neo4j, shares their experience importing GEDCOM files into Neo4j to explore genealogy and ancestry data as a graph. They used a tool that queries GEDCOM files with a syntax similar to JSON, allowing them to extract individual attributes and relationships. The author then used the Python library `python-gedcom` to parse and query the GEDCOM file, creating a dataset for the British Royals and US Presidents. They pre-processed the data by extracting specific attributes and filtering out individuals without parents of the current type. The authors then imported the data into Neo4j using the Neo4j AuraDB Free instance, creating nodes for each person with their attributes and relationships. To add more complexity to the graph, they added inferred relationships such as family ties, sibling relationships, and ancestral relationships. They also explored visualization options in Neo4j Browser and Bloom, discovering a beautiful graph visualization of the ancestry data. The authors provide code examples using Python and Cypher queries to create the data model and import it into Neo4j, highlighting potential issues with uniqueness checks due to non-unique IDs within GEDCOM files.

I, Plushcap, have summarized the text as follows: Harry Corfield, a Data Architect at Telstra, developed his first Neo4j graph database to better understand their data landscape. He started by learning from the Neo4j GraphAcademy and installing the Neo4j Desktop for development purposes. After completing courses on importing data and querying with Cypher, he created nodes and relationships using the LOAD CSV command and performed Cypher queries on his graph database. With the help of additional resources like documentation, community forums, and Stack Overflow, Harry successfully developed his graph database within a two-week sprint. His project earned recognition from Telstra and was even nominated for the annual GraphConnect Awards.

An embedded database is a database used inside another company's application, providing added value and functionality. Graph technology allows organizations to view a single person's information and connect that individual to various common elements, making it easier to spot patterns indicating fraud or suspicious behavior. This technology can be used in various scenarios such as network and IT operations, cybersecurity, retail, and next-generation identity and access management. It enables real-time recommendations, improves customer experiences, and provides a 360° view of master data. Graph databases like Neo4j are well-suited for these use cases due to their ability to create real-time data connections across disparate data sources, analyze complex connections at scale, and provide an interconnected view of relationships.

The author of this article is a Customer Success Engineer at Neo4j, who demonstrates how to integrate BambooHR with Neo4j using its Cypher query language and APOC procedures. The integration allows for the extraction of employee data from BambooHR, which can then be visualized in Neo4j Browser or Neo4j Bloom, creating an organization chart layout. The author also provides examples of queries that can be run on the integrated data, such as finding employees working in a specific department or those at most two relations from the Chief Executive Officer. This integration enables organizations to leverage Neo4j's graph database capabilities for more effective management and analysis of their employee data.

Through its company culture values like "lagom" (harmony and balance) and "positive intent", Neo4j fosters strong relationships among employees, partners, customers, and even with itself, leading to improved work-life harmony, trust, and collaboration. By prioritizing these connections, the company has seen significant growth in teams and partnerships, with a focus on nurturing relationships that extend beyond its Graph Data Platform. This approach has had a profound impact on the author's personal life as well, allowing them to appreciate the value of human connections and their role in technology-driven endeavors.