September 2015 Summaries

In 2008, I joined MarkLogic and became a big advocate for NoSQL technologies. Many problems are better addressed with NoSQL technologies like XML, key-value pair, document, and graph databases. When I joined FactGem as Chief Technology Officer in 2013, we needed a database that didn't require a fixed schema to store user-uploaded data from Excel files without technical expertise. We initially built on MarkLogic, but later realized the importance of data relationships between entities, leading us to transition to RDF-based databases like Neo4j, which met our high performance and flexibility needs, allowing us to focus on relationships between entities in the database.

Neo4j, a graph database, offers powerful tools to manage dynamic relationships between data. Relationships can be used as pointers to identify groups of connected nodes and are also intended to store specific states for sub-graphs. The use case of a linked list demonstrates how consecutive queries can move the relationships between nodes, emulating pointer movement. A three-way switch analogy is introduced to explain combined states, where relationships represent electrical wires, current states of switches (pointer), and circular linked lists around endpoints for each switch. The example shows how these relationships can be used to manage a home automation system, including lamps, doors, gates, or shutters, with potential applications in facility management.

Neo4j is a powerful graph database that can be optimized for better performance. The article discusses various query tuning tips to improve the speed and efficiency of Cypher queries, including hardware considerations, configuration settings, index usage, pattern optimization, defer property access, fast relationship counting, and reducing cardinality of work in progress. By applying these tips, users can significantly reduce database hits, improve query execution time, and increase overall system performance.

The text discusses the integration of transactions in Neo4jClient, a C# library for interacting with Neo4j, a graph database. The introduction highlights the importance of integration testing and how Neo4j's transactional capabilities can be leveraged to write clean and reliable tests. The main body explains the new interfaces and methods introduced by Neo4jClient to support transactions, including `ITransaction` and `TransactionScopeOption`. It also provides examples of using these features for integration testing, such as storing entities in Neo4j with transactions. The drawbacks of this approach are mentioned, including the need for a global `GraphClient` instance and potential performance issues if not managed properly. Finally, the text concludes that the integration of transactions into Neo4jClient allows developers to take advantage of Neo4j's capabilities while maintaining the simplicity of the fluent API.

The third milestone release of Neo4j 2.3 is now available for download, offering improved features such as triadic selection for better recommendations, a "Property Must Exist" constraint to enforce data consistency, and enhanced delete operations. The release also includes new powershell scripts for managing Neo4j on Windows, query warnings in the Neo4j Browser to help with query optimization, and additional features like range queries using index seeks and inequality predicates chaining. With this milestone, Neo4j 2.3 is nearly feature complete, and developers are invited to try out the release and provide feedback.

The Neo4jClient, a main .NET client for Neo4j, has been revived after its developer stepped away. The new version, 1.1.0.1, includes significant improvements such as transaction support, use within a TransactionScope, and authentication capabilities. The client is now being managed by Chris, who aims to keep it in sync with Cypher and integrate Bolt serializer. He also plans to add LINQ integration and better interaction with PowerShell and Windows world. Users are encouraged to provide feedback and collaborate on the project through GitHub or email/tweet.

The text appears to be a collection of metadata and content related to the Graph Data Summit, which is a conference focused on graph databases. The content includes videos, slides, and articles about various topics such as Neo4j, graph databases, recommendation engines, and their applications in different industries and use cases. There are also mentions of other graph database platforms like MongoDB and Azure Cloud Services. The metadata provides information about the speakers, topics, and dates associated with each piece of content. Overall, the text suggests that the Graph Data Summit is a platform for sharing knowledge and expertise on graph databases and their applications.

You're invited to GraphConnect San Francisco, a premier event in the graph database community. The conference features keynotes and big announcements from industry leaders, including Neo Technology's Emil Eifrem and Chief Scientist Jim Webber. There are three different learning tracks to choose from, including lightning talks hosted by sponsors and familiar faces from the Neo4j community. After a day of sessions, attendees can unwind with DisConnect, a networking event that offers opportunities to mingle with fellow graph enthusiasts, speakers, and colleagues. The conference is scheduled for October 21, 2015, at Pier 27 in San Francisco.

The article is about Packt Publishing's Neo4j books, offering exclusive discounts to the Neo4j community. The author has written a brief summary of each book to give an idea of what readers can expect from each title, categorized by beginner, intermediate, and advanced levels. The books cover various aspects of Neo4j, including installation, querying, data modeling, performance optimization, and building web applications with Python and Neo4j. To win a free copy of any of the ebooks, readers can use the discount code NEO4J25 or share this article on Twitter with the hashtag #PacktNeo4j.

The author, a Neo4j user, shares their experience with an IoT project called "fold" that uses Neo4j to manage and visualize data from various devices in their home. The project explores the concept of LAN-centric IoT, where devices are managed locally rather than through cloud services. The author's device captures data on temperature changes caused by the AC unit, which is then visualized using charts generated on a Raspberry Pi. The Neo4j database stores and organizes this data using time trees, allowing for efficient organization and analysis of events over time. The project is still in its alpha stage but has potential interest among Neo4j users who work with IoT devices and programming languages like Java and Eclipse.

The technology industry is leveraging spatial analysis and graph tools to uncover insights in business data, particularly outliers that don't follow established patterns. These outliers can represent emerging relationships, negative interactions, or opportunities for early investment. By visualizing relationships on a map, businesses can identify these anomalies and make informed decisions. Spatial analysis can help expose outliers in various contexts, such as communication, economics, or business development, enabling targeted attention and investment to prevent potential issues from escalating.

Predicting information system incidents with Neo4j can help organizations improve their IT supervision framework by detecting and isolating anti-patterns in the graph. A graph topology, such as one built around integration using Talend to load data, Neo4j to model the IT data model, and QlikView dashboard, enables the detection of complex relationships between applications and incidents. By enriching the data model with properties on nodes and relationships, organizations can gain insights into application types, flow types, and bandwidth usage. Graphs also facilitate agile project management by allowing for rapid changes to the data model, which is essential for building proof-of-concept solutions quickly. Additionally, graphs enable accurate predictive modeling by analyzing real-time data from various sources, such as network probes and incident applications. By leveraging graph databases, organizations can identify patterns and anomalies in their IT systems, predict incidents, and take proactive measures to prevent them, ultimately improving overall system reliability and efficiency.

The text discusses importing the full Stack Overflow dataset into Neo4j, a graph database, using Python and Neo4j's CSV import tool. The process involved downloading the dump files, unzipping them, extracting relevant data with a Python script, and then importing it into Neo4j. The entire process took around 80 minutes to complete for the full dataset, but was significantly faster for smaller datasets. After importing the data, indexes were created, and Cypher queries were used to extract insights such as the top users, tags, and answerers. The graph database provided a rich structure for analyzing relationships between users, questions, and answers, as well as the most active answerers for specific tags. The full Stack Overflow dataset was made available on GitHub, along with instructions and scripts for loading it into Neo4j.