September 2016 Summaries

We are pleased to announce that Structr has completed its transition to Cypher and Bolt protocol, allowing for significant performance improvements and enhanced features. The new version, Structr 2.1, is now running on Neo4j 3.0 using the Bolt driver, providing a faster and more efficient experience for users. This update was necessary due to the evolution of the Cypher query language, which presented challenges in maintaining complex queries with multiple indexes. The new implementation uses a single internal index implementation that translates Structr's object-oriented search attribute trees into Cypher statements, taking advantage of advanced features like inheritance-aware queries and geospatial queries. Additionally, this update enables seamless connection to local and remote databases, offering greater flexibility for users.

The Neo4j team is working on integrating their graph database with other data management systems and sources. To achieve this, they are developing a set of components called `neo4j-etl-components`, which can be used to import data from existing databases into a Neo4j instance. The tool uses Extract, Transform, Load (ETL) approaches and can handle various data integration scenarios, including importing metadata from MySQL schemas and applying default mapping rules for CSV exports. The team has developed a set of mapping rules that interpret relational database structures and constraints as equivalent graph database concepts, such as nodes and relationships. These rules are designed to be flexible and adaptable to different domain models, with the goal of enabling users to easily import their existing data into Neo4j. The `neo4j-etl` tool is now available in the Neo4j contrib repository for testing and contribution, with documentation and GitHub resources provided to facilitate its adoption.

Irina Balaur uses Neo4j as part of the eTRIKS project to develop a knowledge management system for translational medicine, leveraging its ability to manage and integrate heterogeneous biomedical data. She chose Neo4j due to its strengths in data integration and inference of new associations among concepts. Since using Neo4j, she has had positive results, including transitioning her initial database with genetic and epigenetic interdependencies into the graph database, and her colleagues have found it helpful for exploring subnetworks and analyzing metabolites. However, she wishes there were more tools to share networks and do data analytics, or an interface that would help clinicians communicate with biologists. She is looking forward to downloading Neo4j 3.0 to see its new features.

The Grasp Theory project is exploring a new way to catalogue and recall documents that are personally relevant, leveraging Neo4j. Having a graph to represent connections between content is powerful, like Google's PageRank algorithm, but doesn't necessarily mean "relevant" in the context of personal relevance. To address this, the team is using Mazerunner, which integrates an existing Neo4j database with Apache Spark and GraphX to generate graph analytics like PageRank, enhancing the relevancy of searches. By adding PageRank values to nodes in Neo4j and re-indexing them into Elasticsearch, users can tweak search results, potentially providing a more personalized experience. The project is ongoing, with plans to import more data, explore additional graph analytics algorithms, and consider "priming" users' brains for better relevance.

Neo4j is a graph database that has been gaining popularity, and its community is actively sharing knowledge and resources. The Neo4j blog features articles, podcasts, videos, slides, and code repositories from various developers, showcasing the versatility of the platform. The "From the Community" section highlights user-submitted content, including posts, stories, and discussions, which can be featured on the blog. To get your post featured, follow Neo4j on Twitter and use the #Neo4j hashtag. Additionally, users can download a free ebook, "The Definitive Guide to Graph Databases for the RDBMS Developer," to learn more about graph databases in conjunction with relational databases.

There is a software dependency hell that many developers will encounter, characterized as a frustrating maze created by conflicting interdependencies between software components. This can be solved using graph visualization tools like KeyLines, which can help developers better understand their project's dependencies. By creating nodes and relationships in a Neo4j database, focusing on three relationship types, developers can visualize the network and gain insights into structure, important nodes, circular dependencies, license compatibility, and new dependency branches. The tool provides features such as graph visualization, shortest paths, and filtering to help users explore and understand their data.

Neo4j, a Swedish company, offers an open learning culture and collaborative environment where employees are encouraged to share their ideas and help each other. The author of the article had mixed impressions initially but was won over by the team's friendly bosses and willingness to answer questions. Throughout the internship, the author learned about Neo4j's product features, competitors, and market category, gaining valuable knowledge that can be applied to future classes and jobs. The company's authentic culture, characterized by a flat hierarchy and emphasis on teamwork, made the author feel welcome and valued. Overall, the author highly recommends Neo4j as a great place to work for someone looking for an excellent internship experience.

The Internet of Things (IoT) is set to revolutionize various aspects of life, from consumer-oriented devices and applications to enterprise sectors. Graph database technology is expected to play a crucial role in handling the vast amounts of geospatial and time-series data generated by IoT devices. Analysts believe that graph databases will enable organizations to make sense of the IoT data deluge, discover new relationships, and analyze patterns to inform maintenance service intervals, optimize store layouts, and improve traffic flows. As the IoT grows, so does the demand for graph technologies, which are expected to propel advancements in IoT networks and applications. The key benefits of using graph databases in IoT include real-time performance, discovering previously unknown or little understood relationships, and leveraging sensor data to better respond to customer and user needs.

The author of this text is attempting to link data from multiple sources, specifically creating relationships between nodes in a graph database, Neo4j. To accomplish this, they utilize the FuzzyWuzzy Python package for string matching and comparison. The goal is to create meaningful connections between entities, such as drug firms and drugs, or legislators and their information, by identifying similarities in their names. The author preprocesses the strings to remove duplicates and non-alphanumeric characters, sorts them, and then applies FuzzyWuzzy's partial ratio and ratio functions to determine the matching rate. They also introduce a modification step to exclude false positives by filtering common words from the strings. Ultimately, they create relationships between nodes using Cypher queries, storing the confidence level of the match as properties for each relationship. The author hopes this post will be helpful to those facing similar challenges in connecting data in graph databases.