April 2018 Summaries

This week in Neo4j has been about exploring the intersection of graph databases and other technologies such as Tensorflow and GraphQL. Featured community member Fabio Lamanna showcased his work on combining Neo4j with Python to analyze migration patterns, while others have shared tips for passing the Neo4j Certification and using APOC spatial functions in conjunction with the Graph Algorithms A* Algorithm. New releases include a blog post on Neo4j graph databases and GraphQL, a Pypher library update with new features, and a Node-RED node for running Cypher queries on Neo4j graphs. Upcoming events include a talk on text network visualization using Neo4j, Python, and Github API, while users are also encouraged to share their own tweets about Neo4j.

The text discusses the various high-performance graph algorithms available in Neo4j, a growing open-source graph database. These algorithms are used to reveal hidden patterns and structures in connected data, enabling understanding, modeling, and prediction of complex dynamics such as resource flow or information transmission. The algorithms cover traversal and pathfinding (e.g., Parallel Breadth-First Search, Depth-First Search), shortest paths (Single-Source Shortest Path, All-Pairs Shortest Path, Minimum Weight Spanning Tree), centrality (PageRank, Degree Centrality, Closeness Centrality, Betweenness Centrality), and community detection (Label Propagation, Strongly Connected, Union-Find / Connected Components, Louvain Modularity, Local Clustering Coefficient). Each algorithm has its own strengths and applications, such as identifying influential nodes in social networks or optimizing network designs. The text concludes by emphasizing the power of graph algorithms in understanding connected data and their practical uses in various domains.

This week in Neo4j covers various topics including graph visualization using Neovis.js, a feature of the community member Norbert Preining who analyzed the Ultimate Debian Database and created a blog series on it. The workshop "Neo4j <3 Pink Programming" was also held by Louise Söderström and Maria Scharin, where participants learned Cypher basics and a sneak peek at the new date data type in Neo4j 3.4. On the podcast, Irene Iriarte Carretero discussed the recipe-similarity problem with Rik, while Will Lyon created a screencast on graph visualization using Neovis.js to find influential nodes and communities of users in the Russian Twitter Troll dataset. Additionally, Jorge Aguilera showed how to use GORM with Neo4j, and Tomaz Bratanic visualized the GoT dataset using various tools, including SpoonJS and 3d-force-graph. The next week's events include a workshop on "The Fragment Network: A Chemistry Recommendation Engine Built Using Neo4j".

The POLE (Person, Object, Location, Event) data model is being leveraged by practitioners to work with crime data, and it's a great fit for graph database technology and graph algorithms. A proof of concept using publicly available datasets showed that authorities can gain useful insights into investigations by analyzing connections in complex data. Graph technology excels at mining connected data, which is hard to capture through conventional database technologies like RDBMS. By building a Neo4j graph database of 29,000 crimes and generating 106,000 relationships, practitioners were able to find deep and complex networks of connections that suggested obscure family relationships, social associations, and clusters of people and crimes. These insights could support ongoing criminal investigations or initiate new ones, enabling data-driven decision making and maximizing police resources in the face of budget constraints.

Fortunately, graph algorithms are up to the challenge. In this series on graph algorithms, we discussed the value of graph algorithms and what they can do for you. We explored how data connections drive future discoveries and took a closer look at Neo4j's Graph Analytics platform and put its performance to the test. Neo4j offers a reliable and performant native-graph platform that reveals the value and maintains the integrity of connected data. The platform includes advanced graph analytics tools, a growing library of high-performance graph algorithms, and is highly extensible. These algorithms can be called upon as procedures and are customizable through a common graph API. Graph projection is also an extremely handy feature that allows for temporary sub-graphs to be placed into an algorithm when the original graph has the wrong shape or granularity. Neo4j's optimized algorithms yield results up to three times faster than Apache Spark GraphX for certain queries, and can analyze billions of relationships using common equipment in seconds to minutes. The platform is designed to keep computing costs and time investment to a minimum, and offers a growing library of graph algorithms available via the Neo4j Graph Platform.

The article explores how to create graph data visualizations using Neo4j and Neovis.js, a JavaScript library for graph visualization. The goal is to leverage graph algorithms like PageRank and community detection to style the visualization. The author uses the Russian Twitter Trolls sandbox dataset to demonstrate this process, running graph algorithms on the retweet network to find important users and communities. The resulting visualization is then created using Neovis.js, with the pagerank property used to determine node size, community for color, and the count relationship property for relationship thickness. The article aims to provide a practical example of how graph visualizations can be enhanced with graph algorithms, making insights in the data more accessible.

This week in Neo4j saw the launch of several new initiatives and updates, including the long-awaited Graph Gear Store where customers can order Neo4j swag. The community was also active, with Martin Preusse being featured as a community member for his work promoting the use of graphs in life sciences. The company's developer relations team launched a survey to gather feedback from developers, which received positive responses. Additionally, the team published articles on Graph Tour 2018 and graph visualization, and introduced new plugins such as neo4j-graphql-js. Other notable mentions included an interview with Niklas Saers about Swift and iOS development, and several upcoming events including DataScienceFest in Italy.

The Neo4j developer relations team is launching a Medium publication to share lessons learned, tips and tricks, and encourage contributors to share their own content. Those interested in contributing can reach out to the devrel team at `[email protected]`. The publication will cover various topics related to Neo4j and graph databases, with a focus on user feedback to determine the types of articles that are most needed. Readers can follow the official Neo4j Medium account and stay up-to-date with weekly updates from the community through "This Week in Neo4j" posts. The team also invites readers to join the neo4j-users Slack channel for quick questions and technical discussions.

The text discusses hands-on graph data visualization using the Neo4j graph database. It introduces querying data from Neo4j using Cypher, a pictorial graph query language, and provides examples of simple queries to retrieve node and relationship information. The article also covers loading data with the Neo4j JavaScript driver, which is performed efficiently and can handle large amounts of data. Additionally, it mentions various JavaScript graph visualization frameworks that can be used to render graphs, including 3D-force-graph, yWorks yFiles, Linkurio.us OGMA, Keylines, Tom Sawyer Perspectives, Graphistry, and Graphileon. The article concludes by highlighting the importance of hands-on experience with graph data visualization and invites readers to explore these topics further.

SemSpect is a tool that enables users to visualize and interactively query large graphs to gain meaningful insights, particularly in Neo4j property graph data. It groups nodes by their label and aggregates relationships between groups unless the user asks for details, making it easier to explore complex queries without learning Cypher syntax. The tool provides a Web UI based on HTML5/JavaScript with a Java backend that incorporates GraphScale technology, allowing it to draw on full RDFS and OWL 2 RL reasoning capabilities. SemSpect is now available as a Neo4j Graph App for one-click installation and instant use within Neo4j Desktop, providing an effective solution for business-critical graphs with data-driven exploration and visualization.

The presentation discusses the application of graph technology in human capital management (HCM), specifically using Neo4j. The speaker explains how HCM has evolved to focus on people as capital for the company, and how teams form and reform depending on current needs. Graphs can help find hidden potential in organizations and address HCM trends such as people analytics and viewing the organization as networks of teams. The presentation highlights four key areas: recruiting and onboarding, learning, performance evaluations, and talent management. It also touches on finding hidden potential, people analytics, networks of teams, organizational network analysis, and bringing it all together with graphs. Throughout the presentation, the speaker emphasizes the importance of relationships in HCM and how graph technology can help extract value from these relationships.

Graph algorithms offer a powerful approach to graph analytics, utilizing the connections between data to evaluate and infer complex systems' organization and dynamics. They enable data scientists to surface valuable information hidden in connected data, iterate prototypes, and test hypotheses. Graph algorithms are essential for real-time analysis of transactions and operational decisions, providing a local view of relationships between specific data points, as well as global graph algorithms that offer a broad view of patterns and structures across all data and relationships. The best graph algorithms should be easy to use, fast to execute, and produce powerful results, with optimized models supporting different use cases such as real-time recommendations and finding patterns in large datasets. Graph analytics tools must balance performance, scalability, and data integrity, using state-of-the-art algorithms that avoid stalling or recursive processes, and providing trustworthy discoveries through ongoing educational material.

The Neo4j team organized a unique event, called the Community GraphTour, which took place in eight cities across Europe and the Middle East. The community came together to organize their own events, with over 60 submissions from individuals worldwide. Despite the team's guidance and support, the organization of each event was managed by the community members themselves. The Neo4j team is grateful for the enthusiasm and effort of their community members, who made the Community GraphTour a success. The event showcased the power of graph technology to bring people together and drive positive change.

Neo4j, Inc. announced the introduction of a new, highly efficient Cypher query optimizer named Der Hunger as part of its Neo4j Graph Platform, claiming it to be 10,000 times faster than previous optimizers like Ronja. The development of this optimizer, rooted in a heuristic approach dating back to 2008, aims to enhance query planning and execution speed across Neo4j instances. Der Hunger, described humorously as never sleeping and integrated with AI-like algorithms, is set to be released in a beta version to both enterprise and community users, with an unconventional business model involving coffee and sushi as maintenance fees. The announcement, however, is revealed to be an April Fool's joke, emphasizing the continuous improvement of Neo4j's actual Cypher optimizer and encouraging users to seek support through official channels.