May 2016 Summaries

In a German toy manufacturer, Schleich, data management was a challenge due to product conformity and chemical safety concerns. The company implemented a semantic product data management (PDM) system to address these issues. The system uses graph databases, such as Neo4j, to manage complex relationships between products, materials, and regulations. It allows for cross-company data collection and provides a transparent system with high data quality. The system is scalable, accessible, and collaborative, making it suitable for global supply chains. It eliminates the need for separate document management systems and provides context-based access to relevant data. The system also enables easy tracking of product launches, approvals, and regulatory compliance. By using this concept, companies can improve their data management and reduce the risk of non-compliance with regulations.

Neo4j is a popular graph database that has been used by Jim Weaver for his project, which combines Wikidata and Wikipedia to create a semantic navigation system. The use of Neo4j allows for fast querying and storage of large amounts of data, with 11 million nodes and 74 million relationships. Weaver attributes the choice of Neo4j to its Java compatibility and performance, as well as its growing popularity among developers. He has found Neo4j to be a powerful tool for solving real-world problems, with impressive speed and simplicity in querying and importing data. If he could go back in time, Weaver would take advantage of the latest features and tools available in Neo4j 3.0 to streamline his project's data import process.

The International Consortium of Investigative Journalists used Neo4j to analyze connections in the Panama Papers. The ICIJ initially released CSV files, which can be loaded into Neo4j using tools like `neo4j-import` or on the command-line interface. Additionally, a bundled Neo4j database with the data is now available for download from their website. A video demonstrates how to get started with the Mac distribution, and a Docker image is also available for running the investigation. The ICIJ has released more information about their use of Neo4j in the Panama Papers investigation, which can be found on their website. Those new to Neo4j can access a free copy of the Learning Neo4j ebook to get started with the graph database.

The Internet of Things (IoT) is a reality that's already happening, with devices and machines interacting with each other in complex ways. With cheaper chips and decreasing power requirements, IoT devices are becoming ubiquitous, and the value of connectivity is being recognized as a key aspect of business strategy. Examples like Uber, smart buildings, and precision farming demonstrate how IoT data can be used to create valuable processes that couldn't have been created before. Graph databases, complex event processing, and cognitive reckoning are key technologies that enable the Analysis of Things (AoT), which is the value creation of IoT. The future of IoT depends on graph technology, and it's not just about connecting devices, but also about blending IoT data with existing data to extract business value.

Aseem Kishore, a developer at FiftyThree, a New York-based startup, shares his experience with Neo4j, a graph database. He discusses the importance of consistency in reading and writing queries, particularly in handling transient errors and ensuring strong consistency. Aseem explains how he implemented Per-User Read-After-Write Consistency by tracking the last write timestamp and fetching data from the master node for strong consistency. He also addresses issues like race conditions, double-check blocking, and transactional queries to ensure data integrity. The article highlights the importance of considering multiple factors when designing a graph database system, including HA configurations, pull interval, push factor, strategy, and contention awareness. Aseem concludes by emphasizing the need for a balance between simplicity and caution in query design, highlighting his company's approach to abstracting complexity and tackling specific issues as they arise.

Neo4j is being used by Hästens, a company that offers high-quality mattresses, to improve its customer data management and sales force capabilities. The company is leveraging Neo4j's graph database capabilities to create a 360-degree view of its clients, connecting it with SAP systems and sales force to enhance Master Data Management (MDM). By using Neo4j, Hästens has seen faster development processes and unexpected benefits, such as discovering new ways to utilize the graph database. The company is now fully committed to Neo4j, having initially planned for a half-year implementation period, but instead accelerated its adoption due to the positive outcomes they've experienced.

The Neo4j Knowledge Base is a new resource providing permanent and up-to-date links for Neo4j users, covering areas such as Cypher tuning, import issues, and network and hardware setups. The knowledge base uses AsciiDoc with AsciiDoctor, GraphGists, and in-browser guides to make updates easy and quick. It allows users to contribute by raising an issue on the GitHub repository or sending an email to [email protected], with a recommended template for adding content. The knowledge base aims to provide a supplement to the Neo4j documentation, while also being easily accessible from other platforms such as StackOverflow, Slack, and ZenDesk help center.

The text discusses the `UNION` clause in Cypher, a query language used in Neo4j graph databases. The author highlights the issue of applying post-processing operations like sorting and pagination to the results of multiple queries combined using `UNION`. To resolve this, the author proposes rewriting the query using the `COLLECT` function and the `UNWIND` clause. By collecting values into a list with `COLLECT`, then unwinding it back into individual rows with `UNWIND`, and finally deconstructing the maps into columns again with `WITH`, the author demonstrates how to perform post-processing operations on the combined results of multiple queries. This approach allows for more flexibility and efficiency in processing the data, enabling sorting, pagination, filtering, and other aggregate functions to be applied to the results.

Stop Fraud Rings in Their Tracks with Graph Databases [Infographic]` First-party bank fraud and insurance fraud cost billions of dollars annually, while ecommerce fraud rings rack up nearly $4 billion in fraudulent charges each year. The stakes are high, but traditional fraud analytics often overlook sophisticated and organized fraud ring activity that appears normal on the surface. However, graph databases like Neo4j can help detect and prevent such behaviors in real-time by analyzing connections between individuals, devices, and transactions. This infographic highlights the power of graph databases for real-time fraud detection, and a white paper provides more information on how to harness this technology for effective fraud prevention.

Missed GraphConnect Europe this year? You can watch videos of all the sessions right here (with more being added soon!) or check out slides from the presentations on SlideShare. If you would like to see your post featured in May’s “From the Community” blog post, follow us on Twitter and use the #Neo4j hashtag for your chance to get picked. Various articles, including "Immediate Detection of Fraud Rings" and "Panama Papers: How Linkurious enables ICIJ to investigate the massive Mossack Fonseca leaks", were published about the Panama Papers scandal and its connection to graph databases. Several podcasts and interviews with industry experts like Nicolas Rouyer and Ben Nussbaum discuss Neo4j's role in data analysis and graph databases. Videos showcasing Neo4j's capabilities, such as "Faceted search using Neo4j and Graphileon InterActor" and "Create a Recommendation Microservice with Symfony, Neo4j and Reco4PHP", are available on the website. Slides and presentations from GraphConnect Europe 2016 can be viewed online, including "How the power of graphs helps deliver Packt’s strategic vision". Various libraries, graphGists, and code repositories for Neo4j were published, such as "neo4j-php-client" and "node-neo4j: Neo4j graph database driver (REST API client) for Node.js". An ebook, The Definitive Guide to Graph Databases for the RDBMS Developer, is available for download to learn more about how relational databases compare to their graph counterparts.

The International Consortium of Investigative Journalists (ICIJ) is a network of around 200 journalists in more than 65 countries that collaborate on cross-border investigations and issues of global concern. The ICIJ uses technology such as Neo4j, a graph database, to help tell great stories. In the Panama Papers investigation, the ICIJ used open source tools like Oxwall, Blacklight, Solr, Tesseract, Talend, and Linkurious to process 2.6 terabytes of leaked data from Mossack Fonseca, a Panamanian law firm. The team built a private social network for reporters, a mini-Google-like search engine, and visualized the data using graph databases like Neo4j and Linkurious. These tools allowed the ICIJ to cater to both tech-savvy and non-tech-savvy journalists, share information securely, and collaborate on complex investigations. The investigation exposed how offshore tax havens are used by elites from around the world, revealing 140 politicians in over 50 countries with connections to Mossack Fonseca.

The Game Discovery system uses a graph database, specifically Neo4j, to provide personalized video game recommendations. The system's data structure is based on nodes and relationships between them, where each node represents a game or characteristic, and the relationships represent connections between them. The recommendation algorithms rely heavily on the graph layout of the data, allowing for efficient querying and retrieval of relevant results. The system can be queried by searching for specific characteristics or games, and it uses techniques such as shortest path algorithms to determine relevance. By grouping relationships and summing their relevances, the system can efficiently process complex queries and provide high-quality recommendations.

We explored the journey of a small company called FactGem as they transitioned to Neo4j, a native graph database. They initially started with relational databases but realized their limitations and moved on to explore other options like XML databases and Resource Description Framework (RDF). However, RDF's verbosity and performance issues led them to consider alternative solutions. The team eventually found Neo4j, which offered scalability, ease of use, and a powerful graph database that enabled them to build new things quickly. They learned various best practices for using Neo4j, including understanding their strengths, querying techniques, data modeling, optimizing performance, communicating with the community, and keeping code in their toolbox. These lessons helped them navigate the complexities of graph databases and achieve success with Neo4j.

The Neo4j community has shared a wide range of articles, blog posts, podcasts, videos, slides, and code repositories showcasing various applications and use cases for the graph database. The featured content includes tutorials on using Neo4j with different programming languages, such as Java, R, and Python; examples of building recommender systems, fraud detection tools, and network analysis platforms; as well as presentations on topics like data lakes, enterprise architecture, and innovation engines. Some notable resources include the Neo4j Browser's BeerGraphGuide, Graph Karaoke Salute to Anders, and a collection of slides from popular conferences such as SXSW and Qcon 2016. The community is also highlighting various libraries and code repositories for building Neo4j applications, including the neo4j-jdbc driver and the GraphAware Module for Expiring nodes and relationships.