Neo4j Desktop is a really useful application for every graph enthusiast, developer, or analyst who uses Neo4j regularly.

Graphlytic's main goal is to make graph modeling and analytics for day-to-day operations as simple and straightforward as possible. We are constantly adding new features to the visualization and automation modules because we believe that working with graphs, answering questions based on graph data, and task automation with graphs should be easy and accessible even without any or only very little technical knowledge. 

Graphlytic is a graph analytics and visualization web application that can be installed in several ways and one of these ways is to install it in Neo4j Desktop for local usage. This article covers the steps needed to install and run Graphlytic in Neo4j Desktop.

Installation Of Graphlytic In Neo4j Desktop

Supported platforms: macOS, Windows, and Ubuntu (latest versions). Basically, if you are able to run Neo4j Desktop on your machine you should be able to run Graphlytic.

Please contact us at support(at)graphlytic.biz or make a small post to the Neo4j Community portal with any questions or suggestions on how to improve Graphlytic. 

The prerequisite for the installation of Graphlytic in Neo4j Desktop is the installation of Neo4j Desktop. The installation of Neo4j is pretty simple and beginner-friendly. If you haven’t already, download, activate, and start using Neo4j Desktop for free - https://neo4j.com/developer/neo4j-desktop/.

After having Neo4j Desktop installed and activated you can start with Graphlytic:

  1. Open the Neo4j Desktop application on your machine.
  2. Go to the "Graph Applications" page (in the left menu), find the "Install" input field, enter the Graphlytic Desktop app URL: https://npm.graphlytic.biz/graphlytic-desktop and click "Install".
  3. Start a local or remote Neo4j Graph instance - The current version of Graphlytic Desktop works with Neo4j 3.5.6 and higher.
  4. Start Graphlytic by Choosing the "Graphlytic Desktop" application from the "Open" menu in your running Neo4j Graph. If the login screen appears use these credentials:
    • username: admin
    • password: admin
  5. Reindex the full-text index - this is especially needed when you connect Graphlytic to any existing Neo4j graph, with data already loaded in the graph. You will be not able to use the fulltext search until the database is reindexed. More information can be found in the chapter Fulltext Search below the video.

Here is a short clip of all steps in installing and running Graphlytic Desktop with an existing Neo4j Graph:

Next steps and resources

So, what can you do with your freshly installed Graphlytic? There are several use cases where Graphlytic can be very helpful with its features, e.g.:

  • Graph Modeling - manual modeling or graph generated from different data sources.
  • Pattern searching and visualization with simple built-in analytics.
  • Visualization and analysis of graphs with parallel relationships - this is useful particularly for analysis of event logs and communication logs. There is a short video later in the article on this topic.
  • Scheduled Jobs for automatic data update and graph manipulation.

We are striving to get the right balance between two opposite things - simple graph UI and support for complex tasks. We have achieved this by a combination of extensive configuration options and bespoke customization. Graphlytic is ready to be used out of the box for any graph data but the true value is in configuration options like Data Schema, Styling, User Groups, Application Permissions, Data Access Management, or Scheduled Jobs (all described in more detail in the next chapters).

Graphlytic Concepts

 

Fulltext Search

When you want to use the full-text index in Graphlytic, please configure it first and reindex the connected Neo4j graph. This is especially needed when Graphlytic is for the first time connected to an existing Neo4j graph - without reindexing the full-text search will not work (only the first 10 nodes will be accessible). After that Graphlytic will automatically reindex any changes done in the graph. 

Fulltext index configuration is accessible from these pages:

  • "Search & Manage Data" - available in the main menu in the right part of the header
  • "Visualization" - the full-text configuration button is located in the header right next to the search input field

Step for reindexing the full-text index (see picture below):

  1. Open the "Fulltext search configuration"
  2. Choose properties that will be indexed
  3. Click on "Start indexing"

Fulltext search configuration menu

Data Schema

For modeling use cases (manual gathering of data) Data Schema configuration can be used for defining restrictions like read-only properties, predefined lists of values for data inputting which leads to better data quality. Default virtual properties can be also defined in Data Schema. Virtual properties are small JavaScript functions that return a value and are very useful for calculating derived values like performance indicators, flags (true/false values) which can be then used for visualization styling. The nice thing about virtual properties is that they are calculated in the scope of the visualization so when you add or remove nodes and relationships from visualization virtual properties are recalculated and new patterns emerge.

Visualization, Style Mappers and Views

User can modify pretty much any aspect of the visualization with the UI but in most cases, a common understanding and interpretation of the graph data are in place. This common interpretation can be used to create default (globally accessible for all users) styling objects like mappers and default visualization settings which are then used as a default setting every time the user creates a new visualization. 

Search Tabs

With this configuration, it's possible to create a repository of predefined views (queries). These views are then accessible for users on the Search page in the form of a tab that the user can add from the repository with one click. There are two types of these views: query builder which returns data in the form of a paginated table with sortable columns and cypher query where the user inputs any cypher query and visualize the result.

Users, Groups and Application Permissions

Graphlytic is a web application where only defined users have access. Users can be grouped into groups () and these groups can have all sorts of things defined with Application Permission, like if users of this group can only read data or if they can also input data. If they can export data, share visualizations, change global settings, create jobs, and more.

Data Access Management (or Data Security)

Every user group can have different permissions regarding which part of the graph (nodes and relationship) and which properties can be or can not be accessed by users of this group. This allows creating specialized user groups that have restricted privileges like can access the graph itself but can not access financial data stored in properties etc.

Scheduled Jobs

Graphlytic contains an ETL module (Extract, Transform, Load) which allows creating jobs in form of an XML document that defines a set of steps that are executed when the job is started (manually or scheduled with CRON like expressions). Every step can produce a dataset that is then passed as an input to the next step. Steps are defined as the usage of a driver with specific parameters. Graphlytic includes for instance drivers for CSV, Neo4j connection and Cypher execution, Mail, Groovy, Log, Text, XPath. New drivers can be inserted into Graphlytic installation, like when you need a specific JDBC driver or when you want to create your own driver in Java for some use case-specific post-processing after data update.

 

Did Graphlytic get your attention? Please contact us at info@graphlytic.biz if you'd like to ask us anything or to help you assess the viability of the solution for your use case.

 

To understand better what can be achieved with the application take a look at these blogposts:

 

Manual Graph Modeling - see how easy is to use Graphlytic for creating nodes and relationships on the spot.

 

Graphlytic in the Parallel Relationships Graph Models - event analysis, communication, or process analysis can be done very easily with the Parallel Relationships models together with Graphlytic's features like Virtual Properties and Timeline.