November 2016 Summaries

Neo4j has been a passion project for the author, who is a Neo4j consultant at GraphAware, for the last four years. The graph database was used to provide recommendations for holiday house accommodations in just two weeks, and it provided flexibility to make changes over time, which was essential for clients who often weren't sure what they wanted. The author uses natural language processing and Elasticsearch integration with Neo4j to provide graph-aided searches and has been able to go into production much faster than with traditional relational databases. They also use Apache Spark computations within Cypher, allowing the Neo4j JVM to remain quiet while performing computations outside of the main database. Overall, the author is impressed with Neo4j's capabilities and flexibility, which have helped their company increase their customer base.

The LARUS company has been a Neo4j partner since 2015, providing consultancy and training services. They have worked with various companies in Italy, including Cerved, which uses Neo4j to store its business network. The Couchbase connector is an example of the tool's effectiveness, featuring bi-directional relationships and configurable connectors. The architecture of the connector consists of four main components: a Couchbase listener, a JSON loader, a Neo4j mutation listener, and a Couchbase JSON loader. These components work together to transform JSON data into a graph database. The domain-driven transformer allows for reusing "U2" and other nodes, enabling relationships to be built on top of the data. LARUS also contributed to the development of community drivers for Neo4j 3.0, including a standard JDBC interface and a Bolt protocol implementation.

This blog series explores how Neo4j and Oracle RDBMS can work together to deliver optimal results for enterprise architects and business teams. Migrating or syncing a subset of data from Oracle to Neo4j can be beneficial, especially when dealing with complex relationships that cannot be efficiently answered by the current application. Two approaches are discussed: migrating relevant data to Neo4j as a transactional ACID-compliant data store, and synchronizing a portion of the data with Neo4j using middleware such as Oracle GoldenGate. This approach has been successfully implemented by Norwegian telecom Telenor, who synced their resource authorization data with Neo4j, enabling real-time login capabilities for users. The benefits of this approach include faster login times, reduced latency, and improved scalability.

David Meza, an engineer at NASA, uses Neo4j to analyze and visualize data relationships between different projects, such as the Apollo and Orion missions. He chose Neo4j due to its ability to connect data across various spectrums and make it easy for users to visualize and explore complex relationships. Using Neo4j has allowed David's team to save time and money on recent projects, uncover trends in unstructured data, and connect siloed repositories. If he could go back to his first project with Neo4j, he would have better understood the data model and shifted his thinking to focus on relationships and connections. He praises Neo4j for its scalability and capability of handling large amounts of nodes, making it a valuable tool for NASA's "big data" challenges.

The text appears to be from the official Neo4j blog, featuring various articles, podcasts, videos, and other content related to graph databases. The "From the Community" section highlights user-submitted posts, including Twitter analysis of the US Presidential Election 2016, user-defined functions in Neo4j 3.1.0-M10, and more. Additionally, there are resources available for developers, such as tutorials, podcasts, videos, libraries, and code repositories. The blog also promotes the Neo4j Certification exam to showcase skills and get certified in less than an hour.

The author of the text is a data scientist from New York who worked on a project using Neo4j and R to develop a food recommendation engine that aimed to improve people's health. The project was inspired by the growing obesity rate in the United States, with a focus on understanding the connection between food consumption and health outcomes. The dataset used was from the National Health and Nutrition Examination Survey conducted by the US Center for Disease Control and Prevention in 2012. The author developed a graph data model using Neo4j that included nodes for people, eating events, foods, and characteristics, allowing for easy exploration of data and generation of high-quality recommendations. The recommendation engine considered three key components: accessibility, nutrition, and pleasure, with a focus on personalization and customization to cater to individual needs. The author demonstrated the effectiveness of the system by identifying individuals who might benefit from healthier eating habits and providing personalized lunch recommendations that met their nutritional and pleasure requirements. The project highlights the potential of Neo4j in developing data-driven solutions for health promotion and nutrition, with opportunities for further refinement and expansion using additional data points and feedback loops.

In today's data-driven world, well-established relational database management systems (RDBMS) like Oracle continue to be a backbone of millions of enterprise applications. However, by harnessing the power of both a graph database like Neo4j and Oracle RDBMS through a polyglot persistence architecture, organizations can tap into the strengths of each technology to deliver improved performance, agility, and scalability. By using Neo4j for unstructured or highly connected data and leveraging Oracle's capabilities for structured data, developers and architects can create applications that use the right database for the right use case, resulting in superior query performance for connected data and a reduced risk of throwing away existing investments in Oracle infrastructure.

The speaker is a data scientist at Monsanto, working on genomic datasets to improve seed products. They were solving the problem of analyzing complex genetic family trees in their dataset, which was becoming increasingly difficult due to the large scale and complexity of the data. They discovered that their dataset naturally fits into a graph structure, making it easy to model and query using Neo4j. The company built a platform around Neo4j, creating an ecosystem with over 120 applications and data scientists, and has seen significant results in terms of efficiency and scalability. The team had to deal with legacy infrastructure and sync problems, but learned valuable lessons about modeling and coding graph data, and would do things differently if they could begin again. The speaker believes that graph analysis and Neo4j have many untapped use cases in the life sciences, particularly in areas like literature mining and family trees.

Neo4j has achieved significant growth in 2016, with over 2.5 million downloads and more than 200 customers, including large enterprises. The company is now cash flow positive after focusing on its core business and customers. Neo4j has received a $36 million series D funding round to further grow the graph technology space, which has seen rapid growth despite competition from other NoSQL databases. The company believes that graphs offer a huge transformative potential for organizations, enabling them to connect data and gain insights. With this funding, Neo4j plans to continue developing its product, building relationships with enterprise customers, and integrating with other data technologies. The graph technology space is expected to converge around a few leaders by 2020, with Neo4j aiming to be one of those leaders, offering a standard part of an enterprise's infrastructure.

The Neo4j team has developed several tools and integrations to facilitate NoSQL polyglot persistence, allowing developers to integrate multiple databases into a single application. These tools include the Neo4j Doc Manager, which automatically synchronizes data between MongoDB and Neo4j; Docker images for deploying multiple databases, including Neo4j and MongoDB; and the Neo4j Cassandra Data Import Tool, which enables importing data from Cassandra into Neo4j while preserving relationships. Additionally, Neo4j 3.0 introduces updates to stored procedures, Bolt, and drivers, making it easier to build integrations and visualizations. The Neo4j team has also developed a number of community-driven tools and libraries, including the APOC library, which provides over 100 user-defined procedures for working with Neo4j data. These tools and integrations enable developers to take advantage of the strengths of different database technologies and build powerful polyglot applications.

The SAVO Group is exploring the use of graph technology, specifically Neo4j, to improve findability in their sales enablement platform. They track customer activity and want to recommend new content to sellers based on what other sellers have used. The group uses a Cypher query language to create a recommendation engine that can provide real-time suggestions. They leverage AWS Managed Services, including Lambda and API Gateway, to deploy the solution quickly and efficiently. The architecture involves an EC2 instance with Neo4j installed, a Lambda function that runs the Cypher query, and an API Gateway that integrates with the Lambda function. The solution is designed to be scalable and secure, with features like encryption and access control. The SAVO Group plans to expand this proof of concept into a fully-fledged content recommendation engine with multiple endpoints for different types of recommendations.

While organizations seek more from their applications, including features, data variety, agility, speed, and innovative capabilities to uncover rich connections in their data. Connected data enables advanced capabilities such as reduced time-to-market, unified data across disparate systems with master data management approaches, and real-time product recommendations that can save millions of dollars by analyzing complex connections to fight financial fraud. A graph database like Neo4j delivers these capabilities without risking Oracle investments, providing a better bottom-line result for enterprise architects and business teams alike. Relational databases like Oracle are optimized for highly structured, predetermined schemas but struggle with navigating multi-layered relationships in real-time. In contrast, graph databases like Neo4j put data relationships first, enabling developers to be more agile and deliver applications faster due to their schema-optional nature, which abstracts data models and validation to the application tier.

Conrad Taylor, a developer, uses Neo4j for his fashionfreelance.com job site project. He was initially struggling with SQL queries and wanted to build an email service that could extract email handles from Neo4j. He chose Neo4j due to its flexibility, efficient Cypher queries, and visualization capabilities in the Neo4j Browser. Using Neo4j has simplified his work, allowing him to iterate on queries more easily. The project is currently focused on building microservices for email filtering and in-browser recommendations. Conrad plans to leverage Neo4j as a recommendation engine and display job seeker and poster information within the website. He feels confident recommending Neo4j for future projects after completing this challenge.

The Graph Devroom is a co-located event at FOSDEM 2017, focusing on graph processing systems. The event will take place on February 4th, 2017, and features talks from experts in the field of graph technology. The organizers are looking for high-quality technical submissions, ranging from principles to practice, on topics such as graph databases, RDF stores, and distributed graph processing frameworks. Submissions should be based on free and open software solutions or free and open APIs. The event aims to provide a platform for developers to learn from experts and present their latest work to an attentive audience. Submissions can be made through the FOSDEM submission system, with a deadline of November 25th, 2016. Accepted presentations will be recorded, and the final schedule will be published on December 11th, 2016.

Nicola Kamila, an IT manager at Veneto Banca in Italy, and her colleague Alberto De Lazzari, discuss their experience with a graph-based tool used to govern the bank's IT infrastructure and architecture. The project aimed to adopt the microservices pattern and build a multichannel architecture, which relied on relationships between various components. They faced challenges such as documentation and governance, impact analysis, and developing automatic feature activation. To address these issues, they initially tried using a relational database but found it was not suitable for tracking or managing relationships. Instead, they discovered the power of graph databases, particularly Neo4j, which provided them with a natural way to represent entities and nodes connected by relationships. With Neo4j, they built a service catalogue that could answer questions about their IT system, automate steps in the process, and provide impact analysis. The tool allowed them to easily draw their data model on paper and integrate it into their existing systems. Ultimately, they found Neo4j to be a powerful tool for managing relationships, providing a whiteboard-friendly interface, and offering a powerful query language.

This project provides a demo application for building a web app, microservice or mobile app using Neo4j. The application is built with Node.js, Express.js, and React.js, and includes a RESTful API that interfaces with the Neo4j database. The application uses a classic Neo4j dataset: the movie database, which includes nodes for Movie, Person, Genre, and Keyword, connected by relationships as described in an image. To get the project running, users need to clone the repository, set up their environment, and follow instructions in the README file. The application provides APIs for importing CSV files into the database and starting the database. Users can start the Node.js backend, React.js front-end, and API using separate terminal tabs or windows. The application also includes a JavaScript Neo4j Bolt driver that handles calls to the database. The driver parses the results of Cypher queries and returns them in a usable form. The application provides routes and models for building home pages, movie detail pages, and person detail pages. Users can fork the repository and hack away, or find ways to improve the template or JavaScript driver by making GitHub issues and submitting pull requests.