On Graph Database Use Cases & Solutions. Q&A with Tara Shankar Jana (TJ)
Q1. Launched in 2017, DZDconnect, the DZD’s knowledge graph built on Neo4j, serves affiliated healthcare and medical professionals. How are knowledge graphs used for this use case?
Diabetes is one of the most widespread diseases in Germany, with about seven million people afflicted by it. DZD uses a Neo4j knowledge graph to map and analyze biomedical information from sources – PubMed literature data, standardized data from animal models, clinical trial data, and data from other life-threatening diseases – to gain new insights and develop effective prevention and treatment measures.
The DZD knowledge graph consists of 1.8 billion nodes and 4.9 billion edges. Instead of manually searching various databases, as was previously the case, all relevant data is mapped holistically in the graph. This increases the speed of queries and reduces the susceptibility to errors when extracting and aggregating data.
Once the data is mapped, the intuitive and exploratory nature of Neo4j Bloom, a graph data visualization tool, allows scientists to easily answer questions like “how many blood samples from patients under 69 do we have?” or “which studies did the samples come from?” and “what parameters were measured?”
To further identify and investigate different subtypes of Type 2 diabetes, researchers use graph algorithms like Community Detection to identify patient clusters that have shared characteristics such as height, weight, genetic defects, and other markers.
This groundbreaking research is only possible with a knowledge graph that employs a Neo4j graph database to bridge data from the various internal and external data silos, graph data science for powerful graph analytics, and Bloom for intuitive exploration and visualization.
For more information, read this case study.
Q2. Can you tell us a bit about the current status of this project?
Currently, DZD is working on bridging the gap between basic research and clinical studies, as well as mapping clinical standards (FHIR, OMOP) to Neo4j. They’re also developing more end-user applications that are made agile with Neo4j.
Q3. The UK Department for Education provides services across England, but a complex services landscape obscured opportunities for cost-cutting. What was the main challenge for this use case?
The Department of Education (DfE) needed a user-centered approach to see how it could best serve the needs of citizens and stakeholders, like educators. DfE delivers critical services across the education sector, for which they depend on a complex IT infrastructure composed of multiple business systems, services, applications, and components. DfE needed visibility across its entire IT infrastructure and all its connections.
The team at DfE started mapping services to help identify opportunities for modernization and cost-cutting, as well as speed up the development and rollout of new services. Creating a central repository of services information, accessible to all users, would aid in faster decision-making and avoid duplication across the landscape. The team’s initial review of the data environment highlighted where potential reuse could be achieved to enable faster and less costly service delivery.
In summary, DfE had a complex data environment that inhibited strategic decision-making.
Q4. How did they solve such a challenge?
DfE needed to model the many relationships across its IT landscape, both physical and logical. They began working with Neo4j and quickly saw that it would support the DfE product roadmap.
The graphs created a shared view of the service landscape, for which DfE found ways to reuse capability, improve usability, and rationalize, delivering better value services for their users. Neo4j helped DfE visualize their data and the opportunities in one place so that they could focus their efforts on streamlining service delivery and operations.
On the implementation side of the solution, the visualization, provided by Neo4j Bloom, was critical, enabling architects, service owners, and product owners to see valuable insights across the landscape. Neo4j Bloom gives graph technology novicesand experts the ability to visually explore and investigate their graph data. Its illustrative, codeless search-to-visualization design makes it the ideal interface for fostering communication between peers and domain specialists. Visualization enabled DfE to prioritize their efforts based on needs and where opportunities to do more for less.
The DfE graph consists of 55K nodes and 60K relationships. The solution needed mission-critical enterprise functionality, like seamless operability and advanced security features. DfE chose Neo4j AuraDB, a fully managed cloud graph database running on GCP (Google Cloud Platform). As a turnkey SaaS product, Neo4j AuraDB provided DfE with lots of flexibility, including a reliable platform, with zero administration, and a fully automated service upgrade and backup. Also given that AuraDB includes the Basic access version of bloom, DfE was able to roll out graph visualization to all key business users. Neo4j’s architectural flexibility helped with data pipelines that kept visualizations refreshed.
For more information, read this case study.
Q5. Why did they choose Neo4j AuraDB specifically?
Neo4j AuraDB is the easiest and fastest way to build graph applications in the cloud. DfE chose AuraDB for a few key reasons.
As a fully managed cloud service, Neo4j AuraDB offers zero administration, where users provision graph databases in minutes, scale on-demand, and manage backups automatically while the service upgrades automatically without any downtime. This lets them focus on their core applications and innovations without ever worrying about managing the database.
AuraDB is designed to be always on, from the ground up, for mission-critical readiness. With a self-healing architecture and a distributed cluster, AuraDB automatically recovers from infrastructure failures without any interruption. Neo4j’s ACID compliance ensures data is safe and consistently stored.
With Neo4j’s enterprise-class security featuring end-to-end encryption, granular role-based access controls, and dedicated infrastructure, DfE protects their most sensitive data and meets their compliance and privacy needs.
Q6. What are the main results and benefits they obtained in using Neo4j AuraDB?
With Neo4j’s help, DfE was able to prioritize tasks, such as identifying cost savings for service teams by rationalizing components. They could see the cost of each service and how they were procured efficiently, which further helped with eliminating duplication and empower the DfE to make better-informed business decisions. Postcode look-up and geospatial mapping got a huge boost, too.
Neo4j changed their strategic thinking and approach in a big way, including link contract, finance, service documentation, resources, and capabilities information for building significant insights to decision-making.
DfE benefited a lot from the rollout of Neo4j AuraDB, leading to broader company adoption with the cybersecurity team at DfE planning to use it to assess the risk profiles of services across the landscape. Further, National Careers Service (NCS) plans to use it as a social platform to connect the change management, service management, and business continuity teams.
Q7. Let’s talk about Neo4j Bloom 2.0. What is it? And what is it useful for?
Neo4j Bloom gives graph technology novices and experts the ability to visually explore and investigate their Neo4j graph data. Neo4j Bloom is an off-the-shelf graph data visualization tool with an intuitive experience meant for self-guided analysis.
With Neo4j Bloom, practitioners can more easily explore their graphs, quickly prototype solutions, and better collaborate across different teams. Its illustrative, codeless search-to-visualization design makes it the ideal interface for fostering communication between peers and domain specialists.
Developers can connect graph data visualization into their existing applications with ease. And business users can view the results of analyses, and interact with specific parts of the graph, for decision and action support.
With the release of Bloom 2.0, we added some key feature updates to make it easier further for our users and customers to interact with graph databases.
Scene Saving is one new feature that saves multiple scenes, preserving all layout and styling details for later use. Similarly, Scene Sharing helps colleagues collaborate by sharing a read-only scene. Histogram allows users to easily apply styling rules and filters for ad-hoc explorations workflows. There are improved SSO capabilities and editing support, as well as many more UI enhancements.
Q8. What is the benefit of integrating graph data visualizations into no-code workflows with Bloom 2.0?
Using Neo4j Bloom, organizations can take advantage of deep links that launch a query in Bloom, presenting relevant information and simply bringing the power of Neo4j and graph databases to more users.
For example, another internal application in use by an organization might present a table of alerts. Developers can leverage deep links to Bloom to allow users to launch a visualization of the relevant part of the graph directly from this internal application. Subject matter experts like fraud analysts, IT security specialists, and others can interact directly with relevant data and use Bloom’s powerful analytic features to explore and analyze the graph – all without needing to be a database expert or write Cypher queries.
Bloom’s search phrase further bolsters no-code workflows by allowing more technical developers to write relevant queries that get stored in Bloom’s perspectives. Subject matter experts and analysts can then easily use these queries, substituting in their own parameters where needed.
Q9. Anything else you wish to add?
I would love to explain the broader graph data platform to the audience. Neo4j is not just a database, but a graph data platform that creates a foundational tech stack for AI and ML applications.
Here are the different components of the platform that helps with building today’s intelligent applications:
Native Graph Database
The foundation of the Neo4j platform delivers enterprise-scale, performance, security, and data integrity for the transaction and analytical workloads.
Data Science and Analytics
Explorative tools, rich algorithm library, and integrated supervised machine learning framework.
Development Tools & Frameworks
Tooling, APIs, query builder, multi-language support for development, admin, modeling, and rapid prototyping needs.
Discovery & Visualization
Code-free querying, data modeling, and exploration tools for data scientists, developers, and analysts.
Graph Query Language Support
Cypher and openCypher see ongoing leadership and standards work (GQL) to establish lingua franca for graphs.
Ecosystem & Integrations
A rich ecosystem of tech and integration partners, as well as ingestion tools (JDBC, Kafka, Spark, BI Tools, etc.) for bulk and streaming needs.
Deploy as-a-Service (AuraDB) or self-hosted within your cloud of choice (AWS, GCP, Azure) via their marketplace or on-premises.
Tara Shankar Jana (TJ), Sr. Director of Product Marketing (Neo4j Graph Database and User Tools)
TJ is a passionate technologist and product marketing leader with 15 years of experience in databases, cloud, and AI and ML technologies. TJ has helped organizations of all sizes with envisioning, developing, and operationalizing innovative solutions by helping them convert data-driven insight into foresight.
DZDconnect: Using connected data to fight diabetes. Diabetologe 17, 780–787 (2021)
Further information about the DZD project “Graphs to Fight Diabetes”: Video (02:04 Min. – in German)
Sponsored by Neo4j