February 2015 Summaries

Graph databases are gaining popularity as a solution to handle data relationships, which is crucial for next-generation applications that require real-time recommendations, graph-based search, and identity management. They have been around for nearly 300 years, with the concept of graphs theory dating back to Leonhard Euler's work on the Seven Bridges puzzle in 1736. Graph databases are easily understood and used daily, similar to how we navigate the London Underground map or follow family trees. These databases power many everyday businesses, including social media sites like Twitter and Facebook, online retailers such as Walmart and eBay, and online dating platforms, analyzing complex relationships to identify potential matches. The fast and real-time nature of graph databases makes them an excellent platform for unlocking business value from data relationships, which is not possible with traditional SQL or most NoSQL databases.

Graph databases are becoming essential tools in discovering, capturing, and making sense of intricate relationships and interdependencies within the Internet of Things (IoT). The value of devices in the IoT lies not only in their individual capabilities but also in the connections between them. Graph databases store and process data by bringing these connections to the fore, providing significant performance and ease-of-use advantages, as well as unparalleled ease in evolving the data model. This enables businesses to identify opportunities for new services and products that make the most of the IoT, such as optimizing supply chains, analyzing network behavior, and detecting failures. By reducing the wash of data to its common denominators – the connections between devices – graph databases provide a natural way to represent these connections, allowing companies to navigate the complexities of the IoT and unlock new business value.

The author of the text, Mark Needham, has been using various NLP libraries to derive topics for his corpus of How I Met Your Mother episodes without success. He then discovered Prismatic's Interest Graph API, which exposes a web service that takes a block of text as input and returns a set of topics and associated scores. The API is trained on millions of articles shared on social media accounts and has been useful in finding new material to read. Mark uses Python to call the API and requests library to make the calls. He first loads data from CSV files containing episode information, sentences related to each episode, and topic scores. After processing this data, he links episodes with topics using Neo4j, a graph database. The resulting graph shows relationships between episodes and topics, including topics in common and seasons where these topics appear.

Neo4J is a popular graph database used by Cisco in their web-based applications, leveraging its advantages such as intuitive data modeling, improved performance, and speed to import data. Unlike relational databases, Neo4J stores data as nodes, relationships, and properties, allowing for easier modeling of complex information. With Cypher Query, Neo4J enables simple paths to find related information, reducing the need for joins and improving data performance. The database also offers various techniques for importing data, including Cypher statements, JDBC projects, and batch-import projects, making it an ideal tool for data migration and organization in high-performance environments.

The Neo4j.rb gem provides two main classes for building Cypher queries: Neo4j::Core::Query and Neo4j::ActiveNode::Query::QueryProxy. QueryProxy is a convenience class that uses information from models to build queries, whereas Core::Query is more flexible but requires explicit syntax. The key to getting the most out of Neo4j.rb is understanding how to move between these two classes in different situations. Methods like `query` and `query_as` can be used to convert QueryProxy objects to Core::Query objects, while methods like `proxy_as` can be used to convert Core::Query objects back to QueryProxy objects. By leveraging these tools, developers can optimize their Cypher queries for performance and create more efficient code.

The task of finding the right employee among active personnel for a vacant position is challenging for organizations of all sizes, particularly in large businesses where the Human Resources department may not be aware of everyone's skills and competencies. A GraphGist provides a graph data model that simulates this problem and outlines possible semi-automated solutions using collaborative and content-based filtering. The solution leverages both approaches to provide better recommendations for candidates satisfying one or more requirements, and can be implemented with the Cypher query language. A simple internal competence management tool based on Neo4j is also provided for everyday tasks such as organizing training and encouraging cross-evaluation among personnel.

The Neo4j team, led by the creator, has always emphasized the value of relationships in data, and this sentiment is now being validated by the market's growing adoption of graph databases. Several well-known companies, including Walmart and HP, have built highly scalable applications on top of the Neo4j graph database, demonstrating its transformative power. The community surrounding Neo4j has grown rapidly, with over 20,000 meetup members and more than 500 events organized in 2015 alone. Analysts predict that the graph market will explode, with Gartner stating that graph analysis is a competitive differentiator for organizations pursuing data-driven operations. Forrester Research predicts that graph databases will be used by over 25% of all enterprises by 2017 and that the market will reach $400 million in annual spend by 2018. The recognition of graph databases has grown significantly over the years, with companies like DataStax recognizing their value and seeking to support them through products like DSE.

David Montag's article discusses the advantages of using graph databases like Neo4j by posing four pivotal questions that help determine whether such a database would be beneficial for a given project. These questions address the evolving nature of data models, the complexity of multiple JOIN operations, the potential need to explore different data relationships, and the intuitive appeal of visualizing data as graphs. Montag suggests that if one answers 'yes' to any of these questions, considering a graph database could be advantageous. He emphasizes that graph databases can simplify complex relationships and support dynamic data needs more efficiently than traditional relational databases. The article promotes a deeper exploration of graph technologies and offers a free O'Reilly ebook to aid in understanding and implementing graph databases.

Neo4j's community had a great start to the year in 2015, with many people contributing to the project and sharing their experiences with it. The company is grateful for all the posts about Neo4j that were shared on StackOverflow, blogs, and other platforms, which showcased various applications of the graph database technology. To learn more about graph databases and how to use them in application development, users can access a free copy of O'Reilly's Graph Databases ebook by clicking on a download link provided.

Neo Technology, the company behind the popular graph database Neo4j, is positioning itself as a leader in enterprise software with significant momentum gained in 2014, dubbed the "Year of the Enterprise". The company expects 2015 to be the "Year of the Use Case" as customers unlock value in their data relationships using the graph database. According to Harvard Business Review, sustainable business advantage can be gained through network effects of business graphs, which Neo Technology's customers are leveraging to gain competitive advantage by creating new products and services that leverage data relationships or reimagine existing applications with data relationships. Graph databases have been recognized as a key technology in this space, with Forrester Research predicting they will reach 25% of enterprises by 2017. Gartner has named Neo Technology a Cool Vendor and included Neo4j in the Magic Quadrant for Operational Database Management Systems. Customers such as Walmart, eBay, and Zephyr Health Inc. are using graph databases to improve various use cases, including master data management, IT configuration management, and fraud detection. The company's graph database stores and processes data by bringing relationships to the fore, providing significant performance and ease-of-use advantages.

The author attended FOSDEM, a conference in Brussels for passionate developers working on Open Source Software. The Twitter Graph Viz was used to visualize connections between attendees and tags about the event. A room in an old building was too small to accommodate everyone interested in graph theory and graph databases, with standing room only for talks. Neo4j had 6 talks at FOSDEM, including a presentation by Max de Marzi on software analytics. The author met local Neo4j meetup members over dinner, discussing graphs and attending another talk by Helene Astier on environment problems. The event was well-organized and thanks were given to the organizers, speakers, and attendees, with invitations to download latest Neo4j Milestone and submit Graph Gist for a challenge.