January 2018 Summaries

Neo4j is being used by retailers to deliver real-time pricing updates across multiple locations, allowing for competitive pricing and optimized profitability. The graph technology enables retailers to understand their micro-markets and optimize product pricing to match availability, improving margins and sales. Marriott International, a hospitality company, successfully implemented Neo4j to drive revenue and competitive differentiation, reducing publishing times by 96% and server capacity costs by 50%. A graph database can effectively represent the interdependencies between complex variables, enabling fast and efficient calculations of prices in real-time. This technology is crucial for retailers to remain competitive in today's market.

This week in Neo4j covers various topics such as Knowledge Graph Search with Elasticsearch and the use of graph databases in digital humanities. Featured community member Eddy Wong, who has been part of the Neo4j community since 2012, is highlighted for his work on graph database events and his experience with Neo4j. Additionally, there are updates on the podcast featuring Konstantin Lutovich, a long-time Neo4j Kernel team member, discussing the new Async API in the Java driver. Other topics include using Pentaho Integration to load data into Neo4j, jQAssistant for enriching scanned code, and upcoming events such as Data Science in Practice: Importing and Visualizing Facebook Data Using Graphs!

The Neo4j engineering team has completed a series of tests in various environments and workloads, sharing the results with users. The tests were conducted on three different server types: AWS Ubuntu instances, dedicated hardware low-end servers, and dedicated hardware medium-level servers. The performance was tested across several workloads, including internal performance testing workloads, store sizes varying between 1-200GB, realistic read and write workloads against real database stores, targeted micro-benchmarks, and very large data imports. The tests were run on the latest patch release of all supported Neo4j versions, including 3.0.12, 3.1.7, 3.2.9, and 3.3.2. The results show negligible performance impacts across different server types and workloads, with some substantial improvements in Neo4j performance from versions 3.1 → 3.2 and from 3.2 → 3.3. The team will continue testing as new OS and firmware patches become available to evaluate their impact on the product's performance.

Retailers face challenges in managing complex supply chains due to a lack of visibility, which can lead to risks such as contamination, fraud, and single points of failure. Graph technology provides a solution by enabling retailers to manage and search large volumes of data with no performance issues, allowing them to detect problems quickly and take proactive measures. A graph database is designed to handle recursive queries and JOINs, making it suitable for complex supply chain management. The use of Neo4j has enabled companies like Transparency-One to build a platform that provides real-time visibility into their supply chains, enabling them to make data-driven decisions and improve operational efficiency. By reimagining their data as a graph, retailers can transform a complex problem into a simple one and pinpoint critical junctures in the supply network, allowing human managers to instantly fix problems and create a sustainable competitive advantage.

This week in Neo4j covers various topics including categorical PageRank using graph algorithms, more on knowledge graphs, and an interview with Jesús Barrasa about Neo4j's new telecoms practice. Featured community member Alberto Perdomo is highlighted for his work as co-founder and CEO of GrapheneDB, a fully managed solution for graph database hosting in Neo4j. Several articles and presentations were shared by experts on topics such as building product catalogs, knowledge graphs, and visualizing PE files. Upcoming events include the Paradise Papers: Investigating and Analyzing Corruption with Graphs and other discussions on graph databases.

Amazon is the uncontested leader in ecommerce delivery, but smaller retailers can use graph technology to take back the lead. Ecommerce delivery service routing requires visibility into inventory and transit networks, as well as support for complex routing queries at scale with fast performance. Graph databases like Neo4j are well-suited for this task due to their ability to handle highly connected data and optimize routes based on various factors such as time of year and product type. A case study of eBay's use of Neo4j shows that the platform was able to overcome scalability challenges and improve delivery times, demonstrating the potential of graph technology in ecommerce delivery.

Meltdown and Spectre are security vulnerabilities affecting almost all modern processors, allowing a malicious program to read data from another program running on the same server. Neo4j is not directly affected by these exploits but can be impacted if deployed on insecure servers or without proper patches. The company has already released patches for its supported operating systems and expects further OS patches and firmware patches to become available over the next weeks and months. Users are advised to apply relevant patches provided by their operating system vendor, testing them before rolling out to production systems. Further research is ongoing into how these vulnerabilities affect Neo4j's security and performance.

This week in Neo4j highlights the latest news and developments from the graph database community. The featured community member is Chris Leishman, who has worked extensively with Neo4j customers and created tools such as the Neo4j C client and libcypher-parser. Knowledge graphs are also a focus, with an article on how to design an architecture that puts a graph layer over a data warehouse or data lake. The Graph Processing Room schedule for FOSDEM has been finalised, and a new version of the Azure template has been released to support the latest version of Neo4j Enterprise. Additionally, there are tips and tricks for installing Neo4j on Windows, analysis tools such as xhprof-analyze, and information on what is a graph database. The week's tweet highlights the capabilities of Cypher queries in accessing multiple graphs and dynamically constructing new ones.

This guide provides an overview of importing the Bitcoin blockchain into a Neo4j graph database. The process involves converting data from one format to another, which can be challenging due to the structure of bitcoin data. Once imported, the graph database allows for analysis that is not possible with traditional SQL databases. The guide covers key concepts such as the blockchain, blocks, transactions, and addresses, providing a foundation for understanding how to import the data into Neo4j. It also includes Cypher queries for inserting blocks and transactions, as well as examples of results that can be obtained from the graph database. While the guide is not exhaustive, it provides a helpful starting point for those looking to perform serious graph analysis on the blockchain using Neo4j.

Retailers face challenges in managing customer experience, requiring real-time control over inventory, payment, and delivery systems. To overcome these challenges, retailers can use graph technology to personalize the online customer experience by serving relevant content based on customer desires, interests, and needs. This involves analyzing customer behavior leading up to a purchase and using that data to guide customers along a more profitable path. Graph technology allows for combining diverse data sources into a personalization engine, enabling real-time responsiveness and improving customer engagement. A case study of a global sporting goods retailer demonstrates the effectiveness of Neo4j in creating a personalized experience by unifying disparate data models and providing access to relevant data in real time. The use of Neo4j enables retailers to build recommendation engines that offer relevant suggestions to online shoppers, ultimately driving increased revenue and customer loyalty.

The Neo4j community has been active over the holidays with various contributions and projects. The featured community member of the week is Iian Neill, who has worked on a Neo4j-based digital humanities project called The Codex, which aims to create an atlas of events, people, and locations from primary sources. The community has also seen new datasets being loaded onto Azure, online learning courses created, and tools like Causal Cluster Quickstart and Docker Compose scripts for spinning up causal clusters. Additionally, there have been interesting discussions on StackOverflow, including a query to find the average number of interactions per character in the Game of Thrones dataset, and a tweet from Gregg Bolinger about playing with Neo4j and Spring Framework. The community is also exploring topics such as family trees, genealogy, and enterprise architecture models using Neo4j.

Agero, an industry leader in roadside assistance, has been utilizing Neo4j to create predictive roadway analytics for drivers and related service companies. The company is leveraging crowdsourced OpenStreetMap (OSM) data to detect changing roadway and driving conditions, analyze dynamic trends, predict potential consequences of those trends, and improve driver safety and the driving experience. Agero's use case involves analyzing hierarchical OSM data structures that consist of nodes, ways, and relationships, which are used to provide real-time driving conditions. The company chose Neo4j due to its Bolt protocol, stored procedures, and plugin capabilities, allowing for easy integration with their existing processes and programming languages. By utilizing Neo4j's graph database model, Agero can efficiently represent complex relationships between customers, service providers, and drivers, enabling data scientists to run analytics on the data from those relationships. The company's experience with Neo4j has provided flexibility in customizing the indexing for their use case and optimizing performance by moving algorithms into stored procedures.

The text discusses data profiling in the context of graph databases, specifically Neo4j. Data profiling is a widely used methodology to analyze the structure, contents, and metadata of a data source. In a graph database like Neo4j, data profiling helps understand anomalies, assess data quality, and discover enterprise metadata. The article provides practical techniques using Cypher, Neo4j's query language, to perform data profiling on the Stack Overflow Questions dataset. It covers various aspects such as database schema analysis, node analysis, relationship analysis, and uses of the APOC library for advanced graph analysis. The text highlights the benefits of storing data in a graph database, enabling powerful analysis of relationships and unearthed connections among individual data elements.