November 2019 Summaries

The text discusses various topics related to graph modeling and Neo4j. Luanne Misquitta explains the importance of "It depends" when it comes to graph modeling advice, highlighting that there is no one-size-fits-all solution. Tomaz Bratanic demonstrates the seed property feature in community algorithms using a Game of Thrones graph, while Alex Woolford analyzes network traffic on his home network. Jesus Barrasa builds a fashion knowledge graph and Jesús Barrasa also releases version 3.5.0.4 of neosemantics, a plugin that enables the use of RDF in Neo4j. Additionally, Rik starts a series of blog posts exploring data from the Carrefour Basket Data Challenge, while Polley Wong is featured as a community member using various technologies including Go, Node, GraphQL, Neo4j, and Hyperledger.

In this week's #GraphCast, Neo4j all-stars Amy Hodler, Amit Chaudhry and Lance Walter discuss the practicalities of adopting a graph technology vendor, highlighting the importance of choosing a graph database that best fits your business needs. The conversation is a great resource for those looking to navigate the graph database selection process, with a recommended supplement being The Graph Technology Buyer's Guide. To stay up-to-date on all Neo4j videos, subscribe to their YouTube channel and catch new content every week.

The opioid epidemic is a complex issue that involves multiple stakeholders, including corrupt pharmacies, smugglers, and middlemen. Neo4j, a graph database company, is working with healthcare companies to crack down on the roots of the epidemic. Graph databases offer value and reliability in addressing this complex problem due to their ability to model relationships between entities and provide context. In the context of the opioid epidemic, graphs can help identify patterns and connections between individuals and organizations, such as doctor-pharmacy collaborations that lead to excessive opioid prescriptions. Neo4j's graph algorithms, including centrality, path analysis, and community detection, can be used to analyze these relationships and identify key players in the epidemic. By leveraging graph databases and algorithms, companies like Neo4j aim to provide insights that can help prevent or mitigate the spread of opioids, ultimately influencing people's behavior and preventing harm.

Graphs are a natural fit for understanding relationships in data, as they were built to incorporate connections and enrich data with value. In nature, you don't get isolated data points; instead, graphs help us understand the fabric of relationships in data. This context is essential for answering questions like how to navigate between points or find efficient routes. Graphs can also add more value by incorporating additional context, such as companies' innovations and user requests. The intersection of responsible AI and graph technology is crucial, particularly in areas like robustness, trustworthiness, and explainability. By using graphs, we can extract relationships and community detection to identify fraud, improve autonomous decision-making, track data lineage, and reveal bias in the data itself. As AI adoption accelerates, it's essential to consider human values and societal implications, as seen in graph technology's increasing traction in various use cases like drug discovery, financial crimes, and predictive maintenance. Practical tips for responsible AI include debiasing data, involving experts, using developer resources, and adding context with knowledge graphs. Ultimately, graphs enable us to address shortcomings in technically based AI systems and human flaws, making them a crucial component of the future of AI.

The week's video from the NODES 2019 conference features Craig Taverner demonstrating a new library for prototyping spatial algorithms in Neo4j, as well as Tomaz Bratanic building a graph analytics pipeline to explore Rome's transport system. The community also celebrates Satoshi Mayumi, this week's featured community member, who has been actively sharing knowledge and passion with the graph community. Additionally, there are releases of APOC and Halin, as well as the launch of Neomap, a Neo4j Desktop application for spatial data. Other notable mentions include Rik playing around with the Colruyt Data Science assignment and Christian Miles showcasing ReGraph, a React toolkit for graph visualization.

GraphAware is a Neo4j consultancy that provides expert advice at all stages of software development projects using or considering Neo4j, and also offers Neo4j training. GraphAware is the world's number one Neo4j consultancy and a strong reseller in some markets. The company builds solutions around Neo4j, such as its flagship GraphAware Hume Platform, an insight engine driving insights from unstructured data with natural language processing. Kyle McNamara, CEO for the Americas at GraphAware, attributes his choice to partner with Neo4j to its reliability, scalability and thought-leading graph database platform. McNamara predicts a future of graphs moving beyond transactional use cases to more analytical ones, enabling businesses to surface insights and gain competitive advantages through knowledge graph completion. With massive interest in graph technology and momentum behind GraphAware, this trend is considered the most exciting by McNamara, who sees an opportunity for hyper-growth in the industry.

Our goal was to use a virtual assistant to give fast recommendations based on customer requests or interests. We combined IBM's Watson Assistant with Neo4j graph database to achieve this. The system works by asking questions in an order determined by edge weights in the graph, updating the graph based on user responses, and making recommendations. This approach allows for interactive and bidirectional dialog, enabling the virtual assistant to provide useful guidance and helping users find the best answer quickly and efficiently. We believe this technology can be applied to various fields such as sales, medical diagnosis, technical support, and more, with graphs tailored to specific topics or themes. By making virtual assistants reactive and adaptive, we can enhance their capabilities and provide valuable advice to users.

The U.S. healthcare system is plagued by rising costs and declining quality of care, with patients often prescribed painkillers as a first line of defense without addressing the root cause of their pain. The opioid crisis has claimed over 47,000 lives in 2017 alone, largely due to overprescription of opioid-class drugs by pharmaceutical companies. To tackle this issue, healthcare data analysis techniques must be adapted to handle the vast and complex datasets involved. Traditional statistical Business Intelligence (BI) approaches often struggle with volume and complexity, but recent innovations in machine learning (ML) and graph algorithms enable more efficient processing of large, connected datasets. By analyzing CMS Open Payments data combined with other complementary datasets, researchers were able to identify high-paying physicians who prescribed opioid-class drugs, including Dr. Chun, whose practices may warrant further investigation due to suspicious payments from pharmaceutical companies and Medicare fraud allegations. The use of Neo4j Graph Platform allows for efficient analysis of the entire dataset without needing to employ sampling or summarize data, revealing insights into the relationships between doctors, pharmacies, and pharmaceutical companies that can inform policy and improve healthcare outcomes.

The Developer Relations team from Neo4j attended the Big Data London Conference and presented a variant of Amy Hodler's talk on responsible AI. They also published videos from the NODES 2019 conference, including Michael Hunger's APOC standard library tutorial, Tomaz Bratanic's blog post on full-text search in Neo4j, and Nathan Smith's exploration of constrained triad dynamics. The team highlighted Dr. David Bader as a featured community member, who is a graph addict and has published over 250 articles on graph thinking. They also mentioned other community members who have been working on various projects, including performance testing using JMeter, Graph Analysis of Software Traces, Causal Clustering, and the Graphistania podcast.

The presentation "Graph Databases Will Change Your Freakin' Life" by Ed Finkler, CTO of GraphStory, is a captivating talk that covers the fundamentals of graph theory and explains why Cypher code and querying a graph database are so powerful. The speaker starts from the basics of graph theory and progresses to explaining the benefits of using graph databases in various applications. Ed Finkler's presentation was found to be entertaining and engaging by many, including one YouTube commenter who stated that it should be everyone's introduction to graph databases.

The presentation discusses GraphQL, an API query language and runtime for building APIs. It introduces GraphQL's key concepts, such as entry points, selection sets, and schema traversal. The presentation also covers Neo4j's integration with GraphQL, including the GraphQL database plugin and neo4j-graphql-js, which provides a JavaScript library for generating Cypher queries from GraphQL schemas. Additionally, the presentation touches on the GRANDstack, a set of tools that includes GraphQL, React, Apollo, and Neo4j, designed to build modern applications. The talk also covers how to use Apollo Client with React to query GraphQL APIs and how to generate Cypher queries using neo4j-graphql-js in other languages.

The author is exploring the importance of responsible AI and its context. They discuss four key issues: myths about AI, biased AI, unknowable AI, and inappropriate AI. The author argues that these concerns are often overlooked or downplayed, but they can have significant consequences, such as perpetuating bias, denying individuals their rights, and infringing on personal freedoms. To address these issues, the author emphasizes the need for context in AI development and deployment, including transparency, fairness, accountability, and public trust. They also highlight the limitations of AI without context, which can lead to narrow focus, subpar predictions, and a lack of transparency. The author suggests that graphs are a natural fit for adding context to AI systems, enabling them to make more informed decisions and improve their performance. Ultimately, responsible AI requires a nuanced understanding of its limitations and the importance of human values in guiding its development and deployment.

The Neo4j community is actively engaged, with various initiatives and resources available. The launch of Neo4j AuraDB, a fully managed Database as a Service, marks an exciting development in the company's offerings. Alicia Frame shares insights on graph embeddings, while Matt Casters provides guidance on writing a Kettle connector. Additionally, David Allen offers a graph modeling master class, covering topics such as relationships and data normalization. The community is also active in exploring structural balance using Neo4j, with Nathan Smith delving into the concept of alliances. Meanwhile, Tiago Oliveira shares his experience completing a course on applied graph algorithms with Neo4j, highlighting the platform's potential for robust Recommendation Systems. The community comes together to support one another, with featured members like Mike Black, who contributes to various aspects of the community, including organizing local events and providing technical guidance.

In this interview, Forrest Swope from UVA discusses how his team is using Neo4j's graph technology to move beyond relational databases and capture complex relationships in their data. They are exploring the challenges faced by community college students who transfer to public universities, with the goal of identifying key factors that contribute to success or struggle. The team believes that graph technology allows them to uncover meaningful relationships between entities, which is essential for solving complex problems like this one. Swope highlights the limitations of relational databases in handling complex relationships and the power of Neo4j's ability to answer questions without prior knowledge of what they will ask.

Neo4j AuraDB is introduced as a fully managed, native graph Database as a Service (DBaaS) designed to provide developers with a seamless cloud-based experience, focusing on accessibility, scalability, and reliability. Built by Neo4j, Inc., AuraDB offers always-on availability with intelligent self-healing features and ensures data durability through ACID transactional consistency and data replication across multiple physical disks. The platform allows for on-demand scalability without downtime, making it ideal for handling peak loads, and supports various developer-friendly integrations such as Cypher, GraphQL, and cloud functions. This service aims to bridge the gap for individual developers and small teams who previously had limited access to the enterprise-grade features of Neo4j, fostering the development of innovative applications with its powerful graph database capabilities.

Money launderers are continually evolving, making it essential for organizations to utilize advanced tools such as graph technology to stay ahead. This technology helps in tracing and tracking complex money transactions, identifying the source of fraudulent activities, and ultimately aiding in compliance with reporting standards. By leveraging graph technology, analysts can efficiently weed through intricate financial networks, freeing up their time to focus on more critical tasks.

Wolfgang Hoeck shares his experience on building a knowledge graph from scratch in a talk at the NODES 2019 conference, while new releases of the Graph Algorithms Library have improved the in-memory graph. Dr. Alicia Frame discusses the application of graphs, AI, and ML, and Nathan Smith solves the market clearing price problem using Neo4j. Chris Farrell has released a tool for querying Bloodhound data, and Arthur Namias de Crasto is featured as this week's community member, who is working on a Neo4j hobby project related to compliance around documents.

This week, I wanted to make sure you didn’t miss a single thing from the Neo4j Online Developer Expo + Summit (NODES) we hosted. With many major announcements and simultaneous talks from more than a dozen time zones, it was easy to miss some of the great presentations, discussions and stories shared at our first-ever online developer conference. CEO & Co-Founder of Neo4j, Emil Eifrem gave a keynote that included the announcement of Neo4j’s forthcoming graph database as a service offering, new projects from Neo4j Labs, preview features from Neo4j 4.0, and other awesome news. We couldn't be more proud of all the speakers who made NODES 2019 a smashing success for the Neo4j community, with many talks now available to watch for free on our YouTube channel.