January 2015 Summaries

Neo4j is a fast-growing graph database used by several prominent companies, including eBay, HP, Cisco, and Walmart, to improve their ability, performance, flexibility, and cost. It helps customers address various challenges such as degrading systems, accommodating new requirements, improving customer experience, and optimizing processes. The database is particularly useful in scenarios where data is hierarchical or complex, allowing for real-time lookups and scalability. By leveraging Neo4j, companies like Telenor, Pitney Bowes, CrunchBase, and TomTom have successfully implemented graph-based solutions that provide them with a competitive edge and improved business outcomes.

Neo4j has been named as a 2015 Technology of the year by InfoWorld, highlighting its growth and success in strengthening its product, community, and personnel. The graph database is known for its scalability, with a learning curve short enough to tackle tasks from small to large, while maintaining ACID-compliance and two-phase commit transaction support. Its API allows modification of shortest path algorithms, and a declarative query language called Cypher provides an alternative to SQL. Neo4j offers both free open source and enterprise editions, with additional features such as clustering, caching, backups, and monitoring capabilities available in the latter. The company's online documentation is highly regarded for its interactive nature, allowing users to explore graph database concepts and execute code against a temporary database.

While SQL is a great tool for storing information, it's not ideal for finding connections among users to enhance a website's social experience. The Flask tutorial builds a microblog application using SQLite, but an expanded version with Neo4j-powered social features provides better results. A graph model allows for easy querying of questions like "What are the top tags of posts I've liked?" or "Which user is most similar to me based on tags we've both posted about?" These queries can be answered using Cypher, a query language used in Neo4j. The expanded application includes features like recommending similar users and displaying similarities between two users when one visits another's profile.

This Winter Challenge aims to encourage developers to create innovative graph models and example queries, focusing on various use cases such as solving complex JOINs, providing recommendations, conducting impact analysis, managing content and assets, and showcasing creative graph search applications. The challenge offers a grand prize of $500 for each of six categories, with the opportunity to participate in the jury for previous winners. Participants have until February 22nd to submit their entries, which will be reviewed and announced shortly after. To aid in the process, Neo4j has launched its GraphGist Portal, featuring a collection of past submissions, and invites developers to share their work using the hashtag #graphgist.

Phone operators can provide valuable information to law enforcement by analyzing phone call data. In this article, we explore how to use graph technologies to analyze phone calls and identify potential suspects in a crime scene. We start by defining a data model that represents the relationships between phone numbers, locations, and people involved in the calls. The data is then imported into Neo4j, a popular graph database, using Cypher, its query language. By querying the graph database, we can find the potential suspect who made a call near a specific location at a certain time. We also explore how to create relationships between suspects and their known associates, which can help identify the perpetrator in a crime scene.

The Neo4j team is hosting the GraphGist Winter Challenge, a competition to come up with innovative graph models and example queries. The challenge focuses on six use cases that graph databases are well-suited for, including solving complex relational joins, generating personalized recommendations, and analyzing impact in dependent information networks. Participants have until February 22nd to submit their entries, which will be judged based on creativity and usefulness. The winners will receive a $500 Amazon gift card, and some past winners will participate as jurors to help select the top entries.

Mix FiftyThree developed an app called Paper that made drawing on the iPad into a complete joy, now with the added feature of Mix, enabling collaboration with others remotely. The new feature allows users to riff on one another's sketches in a richly constructed, annotated system that ensures proper credit is given. This system feels like an evolved Internet that can be reached out and drawn on. The development of Mix was powered by Neo4j, a graph database technology used in the backend of the application at GraphConnect 2014.

The PageRank algorithm, developed by Larry Page and Sergey Brinn in 1996, is a key concept in computer science that determines the popularity of web pages based on the number of links pointing to them. The algorithm has been improved upon by Kenny Bastani, who introduced the concept of Categorical PageRank, which breaks down PageRank into categories or partitions to analyze causality and understand contributing factors to changes in PageRank values over time. Bastani's design uses a property graph data model to partition nodes into pages and categories, allowing for scalable distribution of PageRank jobs using Apache Spark and Neo4j, enabling the analysis of subgraphs that describe each category. The proof of concept demonstrates the feasibility of this approach, and the resulting system can be used to analyze Wikipedia articles and other knowledge graphs.

2014 was a remarkable year for Neo Technology, with over 500,000 downloads of Neo4j 2.0 since its launch in January, thousands of production deployments, and a record turnout at GraphConnect San Francisco 2014. Analysts recognized Neo as a Cool Vendor by Gartner and included it in the 2014 Magic Quadrant for Operational Database Management Systems. The company grew its customer base to over 150 paying subscribers and doubled its revenue. A $20 million Series C funding round was secured, which will enable investments in the product, salesforce expansion, and marketing efforts. Neo Technology has established itself as the clear leader in graph databases, with customers including top companies and exciting startups, and is poised to thrive as the graph market explodes, driven by growing demand for data-driven operations and decisions.

The text discusses various techniques for processing and analyzing text data using graph databases, specifically focusing on word relationships and frequency analysis. It demonstrates how to split text into words, create relationships between consecutive words using graph nodes and edges, and count occurrences of words and their sequential relationships. The document also covers advanced features like recording word positions, handling large files, converting text to lowercase, removing punctuation, trimming whitespace, and filtering out stop words. Additionally, it includes examples of loading text from external files, utilizing periodic commits for large datasets, and querying paths with specific relationship conditions to extract phrases based on their frequency scores.

Neo4j has a powerful ETL tool called LOAD CSV [Neo4j Docs] Cypher clause that can import data from a CSV file into the database. This clause supports loading data from local filesystem or remote URI, performing multiple operations in a single statement, and can be combined with USING PERIODIC COMMIT to group operations into transactions for large amounts of data. To successfully import data, one needs to have the graph data model ready, tune cache and heap configuration, set up indexes and constraints, and use the LOAD CSV statement with common options such as accessing via column header or column index, configuring the terminator character, and using PERIODIC COMMIT to speed up the process. When loading data into GrapheneDB Neo4j instance, one needs to consider specific limitations and recommendations such as configuration restrictions on lower-end plans, lack of authentication support in neo4j-shell, and need for a publicly available URL when running from browser UI.

The release of Neo4j.rb 4.1 and Neo4j-core 4.0 brings several significant improvements, including enhanced performance, support for Neo4j 2.2, improved relationship types, and increased power in scopes. The new features include full support for Neo4j 2.2, allowing users to work directly with authentication REST endpoints, as well as significant speed improvements of over 2x on most repeat queries. Additionally, the release introduces better automatic relationship types, making it easier to manage relationships between models, and all create, update, and destroy actions are now wrapped in transactions for improved performance. The new features also include more powerful scopes, cleaner code, and new association options such as dependent and unique. Users upgrading from Neo4j 3.x can follow a relatively simple process to take advantage of the improvements.

When using neo4j for the first time, most people want to import data from another database to start playing around. To create a simple SQL to Neo4j import process possible, Brian Underwood created neo4apis-activerecord. The setup involves installing required gems and configuring a database.yml file. The command `neo4apis activerecord all_tables --identify-model --import-all-associations` imports data from another database into Neo4j. However, there's one small catch: when referencing foreign keys in the imported data, additional configuration is needed to specify the corresponding table names. This can be achieved by creating a config/environment.rb file and defining the relationships between tables using ActiveRecord's API.

The text discusses visualizing time-based graphs with Neo4j using the KeyLines Time Bar. To achieve this, one needs to load time-stamped data into Neo4j, create a client-side web application, send Cypher queries to retrieve data from Neo4j, parse the response into KeyLines' JSON format, and load the result into KeyLines. The process also involves adding interactivity to the time bar by binding functionality to its 'change' event. This allows the application to query Neo4j for new data when the user extends the date range, enabling real-time visualization of dynamic graphs.

The database industry is shifting its focus towards graph databases, which are increasingly being used by companies like Twitter and Facebook to store relationships and activity between users. Graph databases are based on the mathematics of graph theory and store data in terms of nodes, properties, and edges. These databases are finding applications in social media platforms, online dating sites, and even analytics engines that pull data from various sources. The use of graph databases is becoming more widespread as companies look to leverage their power for relationship-based analysis and recommendation systems.

The Neo4j community continued to produce impressive graph projects despite the holiday vacation period, with various blog posts, presentations, and videos being shared. To have their work featured in a blog post or video series, individuals can share links on social media using the #Neo4j hashtag or tagging @Neo4j. The Neo4j community also released several interesting and innovative graph-related projects last month, including a Graph Karaoke video by Rik Van Bruggen that offers a "bright side" of life with music and graphs. Additionally, individuals can access a free copy of O'Reilly's Graph Databases ebook to learn more about graph databases and how they can be used for applications.