<p><img src="https://user-images.githubusercontent.com/56643819/72578859-51099b00-3912-11ea-8435-3f4d412d004e.png" alt="rc2" /></p>
<p>In our RC2 version, we have made several improvements re the nGQL query language, our Storage Engine, CSV Importer, as well as added Java, Python, and Go Clients.</p>
<p>Below is a detailed rundown of the newly added features:</p>
<h2 id="post-query-engine">Query engine</h2>
<ul>
<li>Support <code>GROUP BY</code> to group items that have the same values, often used with aggregate functions (COUNT, MAX, MIN, SUM, AVG, etc.) to group the result-set by one or more property field. <a href="https://github.com/vesoft-inc/nebula/issues/749">#749</a></li>
<li>UPDATE CONFIGS supports expression  <a href="https://github.com/vesoft-inc/nebula/issues/1273">#1273</a></li>
<li>Support changing log severity level and verbose level from console  <a href="https://github.com/vesoft-inc/nebula/issues/1273">#1273</a></li>
<li>Support IF…RETURN… to return the result if a specified condition is true <a href="https://github.com/vesoft-inc/nebula/issues/1233">#1233</a>&nbsp;、<a href="https://github.com/vesoft-inc/nebula/issues/1246">#1246</a></li>
<li>String properties could have unlimited length  <a href="https://github.com/vesoft-inc/nebula/issues/1103">#1103</a></li>
<li>When insert vertices and edges, users can specify properties order instead of following the schema defined order <a href="https://github.com/vesoft-inc/nebula/issues/1219">#1219</a></li>
<li>tag/edge type properties can have explicit default values.  <a href="https://github.com/vesoft-inc/nebula/issues/860">#860</a></li>
<li>GO FROM … REVERSELY to implement reverse query <a href="https://github.com/vesoft-inc/nebula/issues/1349">#1349</a></li>
</ul>
<h2 id="post-storage-engine">Storage engine</h2>
<ul>
<li>Support SHOW PARTS to fetch the partition information of current space <a href="https://github.com/vesoft-inc/nebula/issues/1086">#1086</a></li>
<li>Support BALANCE STOP to stop load balance process <a href="https://github.com/vesoft-inc/nebula/issues/1238">#1238</a></li>
<li>Support triggering the compaction and flush manually <a href="https://github.com/vesoft-inc/nebula/issues/677">#677</a>, <a href="https://github.com/vesoft-inc/nebula/issues/1240">#1240</a></li>
<li>Change BlockBasedTable default block cache size from 4MB to 1024MB  <a href="https://github.com/vesoft-inc/nebula/issues/1248">#1248</a></li>
<li>Change default partition from 1024 number to 100 <a href="https://github.com/vesoft-inc/nebula/issues/1330">#1330</a></li>
<li>Add max_edge_returned_per_vertex gflags option to limit the return size of supper vertex <a href="https://github.com/vesoft-inc/nebula/issues/1221">#1221</a></li>
<li>Add cache for vertex <a href="https://github.com/vesoft-inc/nebula/issues/1294">#1294</a> <a href="https://github.com/vesoft-inc/nebula/issues/1268">#1268</a></li>
<li>Upgrade dependencies folly and fbthrift <a href="https://github.com/vesoft-inc/nebula/issues/1161">#1161</a></li>
</ul>
<h2 id="post-tools">Tools</h2>
<ul>
<li>Enhance CSV Importer, user can import CSV using docker image. See <a href="https://github.com/vesoft-inc/nebula-importer">Nebula Importer</a> for more details.</li>
</ul>
<h2 id="post-client">Client</h2>
<ul>
<li>Add Python client, see Python Client for more details.</li>
<li>For Golang client, see Go Client.</li>
<li>Java client has been uploaded to Nexus Repository, see Java Client for more details.</li>
</ul>
<h2 id="post-changedremoved">Changed/Removed</h2>
<ul>
<li>Remove commands ADD/REMOVE HOSTS , the hosts could be discovered automatically <a href="https://github.com/vesoft-inc/nebula/issues/1172">#1172</a></li>
</ul>
<h2 id="post-coming-soon">Coming soon</h2>
<p>See our <a href="https://github.com/vesoft-inc/nebula/wiki/Nebula-Graph-Roadmap-RC3">roadmap for RC3</a> for the features coming soon.</p>
<h2 id="post-you-might-also-like">You might also like</h2>
<ol>
<li><a href="https://nebula-graph.io/posts/nebula-graph-rc3-release-note/">NebulaGraph RC3 Release Note</a></li>
<li><a href="https://nebula-graph.io/posts/nebula-graph-rc4-release-note/">NebulaGraph RC4 Release Note</a></li>
</ol>

New features re NebulaGraph’s Query Engine, Storage Engine, and OM tools have come in the RC2 version of NebulaGraph. Check what’s coming.

Features

System-testing

Dev-log

Performance

Use-cases

Community

Deployment

Tools

Query-language

Graph-computing

Tech-talk

Release-notes

Architecture

News

Source-code

Success-stories

Knowledge-graph

NebulaGraph RC2 Release Note

Benchmarking the Mainstream Open Source Distributed Graph Databases at Meituan: NebulaGraph vs Dgraph vs JanusGraph

Graph Database vs Relational Database: What to Choose?

Difference Between Relational and Non Relational Database

Use cases of graph databases in real-time recommendation

We are excited to announce the General Availability of NebulaGraph Enterprise v5.0, a groundbreaking release that establishes [NebulaGraph as the first distributed graph database to offer native support for the Graph Query Language (GQL)](https://www.nebula-graph.io/posts/nebulagraph_enterprise_5.0_gql_support).This milestone represents the pinnacle of three years of focused research and development, showcasing our commitment to innovation and excellence in the graph database industry.

Recognizing the importance of GQL early on, we embarked on this ambitious project in November 2021. The initial phase involved a thorough analysis of the GQL standards, followed by extensive design discussions. Over the next few years, we engaged in multiple development iterations and nearly a year of rigorous validation, continuously refining our implementation as the standards evolved. Today, we are proud to unveil NebulaGraph Enterprise v5.0.

# Why Native GQL Support Matters More Than GQL Compatibility

Implementing GQL at the architecture level is a complex and challenging endeavor, requiring a well-designed architecture and significant enhancements across the metadata, computation, and storage layers of the database. You could refer to the enhanced kernel engine in [Announcing NebulaGraph Enterprise v5.0 RC release: Delivering Full-Fledged GQL Support](https://www.nebula-graph.io/posts/rc_release_of_nebulagraph_enterprise_v5.0).

Some might assume that implementing a new query language is as simple as syntactic translation or adaptation. However, our experience tells us otherwise. Pure syntactic adaptation is fraught with challenges. Existing graph languages, such as Gremlin and Cypher, performing syntax translation to GQL might be in lack of catalog and schema management, which are fundamental to GQL as it draws heavily from SQL standards.

Historical attempts to translate query languages have not been successful. For example, some graph database vendors that have dedicated effort to convert Cypher to its own query languages have seen little progress. This underscores the necessity of native GQL support at the architecture level for achieving optimal performance and full compliance with the standards.

# Key Enhancements in NebulaGraph Enterprise v5.0
## Performance Boosts Up to 3x Faster with 90% Less Memory Overhead

[NebulaGraph Enterprise v5.0 delivers a significant performance boost, achieving up to 3x faster query execution while reducing memory overhead by nearly 90%](https://www.nebula-graph.io/posts/technical_preview_nebulagraph_enterprise_v5.0#post-performance). This is evidenced by the results from the LDBC SNB Interactive benchmark, where the Enterprise v5.0 demonstrated an average speedup of 3.01 times compared to the previous version.

## Graph Projection: Efficient Subgraph Analysis

NebulaGraph Enterprise v5.0 introduces support for graph projection, a powerful feature that allows users to create subgraphs based on existing graphs or MATCH queries. This capability enables users to isolate and focus on specific portions of their graph data, making it more efficient to perform complex graph computations and queries. By working with subgraphs, users can significantly reduce the computational overhead and improve query performance, leading to faster insights and more responsive applications. 

## Multi-Level Permission Management
Security and compliance are paramount in modern data management. NebulaGraph Enterprise v5.0 supports multi-level permission management, offering fine-grained access control down to the property level. This hierarchical permission system spans service groups, schema, graphtype/graph, nodetype/edgetype, and property, ensuring that only authorized users can access sensitive data. This feature not only enhances security and compliance with data privacy regulations but also improves administrative efficiency by reducing the risk of human error.
## Query Optimization and Isolation
NebulaGraph Enterprise v5.0 introduces substantial advancements in query optimization and isolation, including query-level resource limits and index hints. The query-level resource limits allow each query statement to utilize a specified amount of CPU and memory resources, and users can individually set the response time for each query. The major benefit of this capability is that it prevents a few resource-intensive queries from monopolizing system resources and potentially bringing down the entire cluster. This granular control ensures that the system remains stable and responsive, even under heavy workloads. The index hints give users more control over query execution. By specifying which indexes to use, users can optimize query performance and ensure that the most efficient execution path is taken. These improvements are crucial for enterprises handling complex relational data where efficiency, consistency, and security are paramount.

## Upgraded UI Experience for Analysts and DBAs
The new UI experience of visualization tools in NebulaGraph Enterprise v5.0 offers significant improvements for both business analysts and DBAs:
- Explorer: Redesigned for better interaction and focus on analysis, the new Explorer interface allows users to save and load historical analysis results, making it easier to track and revisit previous analyses.
- Dashboard: The Dashboard will integrate alert monitoring and diagnostics, enabling quick root cause analysis. It will also support multi-tenant management and resource allocation, providing DBAs with powerful tools to manage and optimize their database environments.

# Conclusion

NebulaGraph Enterprise v5.0 represents a significant advancement in graph database technology, offering native GQL support to meet the growing demands of complex data environments. We invite you to explore the capabilities of NebulaGraph Enterprise v5.0 and join us in advancing the frontiers of graph database technology. [Reach out to us today!](https://www.nebula-graph.io/contact)


We are thrilled to announce that NebulaGraph Enterprise v5.0 is now GA! This groundbreaking release establishes NebulaGraph as the first distributed graph database to offer native support for GQL.

Announcing the General Availability of NebulaGraph Enterprise v5.0: The First Distributed GQL Database

In the digital age, recommendation systems have become an integral part of online experience, driving better user experience and engagement. A recommendation system, often integrated with ad serving systems, is designed to suggest highly relevant content or advertisements to users based on users’ profile and engagement history. However, the complexity of user interactions and preferences demands robust data management solutions. Enter graph databases—an innovative approach that enhances the efficiency and accuracy of recommendation systems. This blog explores how recommendation systems work and how graph databases empower recommendation systems, providing real-life examples and architectural insights.

# What are Recommendation Systems and How They Work
## Key Components of Recommendation Systems:
- User Data: Information about users, including their demographics, behavior, preferences, and past interactions. This data is crucial for understanding user interests and making personalized recommendations.
- Content Data: Information about the content or ads available for recommendation. This includes metadata, categories, keywords, and performance metrics.
- Recommendation Algorithms: The core of the system, these algorithms analyze user and content data to generate personalized recommendations. Common types include collaborative filtering, content-based filtering, and hybrid methods.
- Real-Time Processing: The ability to process data and generate recommendations in real-time, ensuring that suggestions are timely and relevant.
- Feedback Loop: Mechanisms to collect user feedback on recommendations (e.g., clicks, likes, shares) to continuously improve the accuracy and relevance of the system.

## Types of Recommendation Systems
Recommendation systems are powerful tools designed to analyze user behavior and preferences to deliver personalized content and suggestions. These systems operate by collecting extensive data on user interactions, such as browsing history, ratings, and purchase activities. Using this data, recommendation systems employ sophisticated algorithms to identify patterns and relationships within the data. 

Two common approaches are collaborative filtering and content-based filtering. Collaborative filtering analyzes the behavior and preferences of similar users to make recommendations; for example, if two users have a history of liking similar movies, the system might suggest a movie liked by one user to the other. Content-based filtering, on the other hand, focuses on the attributes of items and recommends items that are similar to those a user has previously enjoyed. By combining these methods, recommendation systems can generate highly personalized suggestions that cater to individual tastes and preferences. This tailored approach not only enhances the user experience by presenting relevant content but also drives higher engagement and satisfaction, ultimately benefiting both users and businesses.

# What are Graph Databases
## Definition and Structure
Graph databases are a type of NoSQL database designed to handle and represent complex relationships between data entities. Unlike traditional relational databases that use tables to store data, graph databases use nodes, edges, and properties to create a flexible and highly interconnected data structure. Nodes represent data entities, edges denote the relationships between these entities, and properties provide additional information about the nodes and edges. This structure allows graph databases to efficiently model, store, and query intricate relationships and dependencies within data. Graph databases are quite young compared with relational databases and well-known graph databases include Neo4j, NebulaGraph, Amazon Neptune, and TigerGraph.

They excel at managing complex relationships because their inherent design closely mirrors real-world networks, such as social connections, organizational hierarchies, and financial transactions. This enables rapid traversal and querying of relationships, providing powerful insights and enabling applications like fraud detection, recommendation systems, and network analysis to operate with high performance and scalability.

## Architectural Insights

1. Data Modeling:
    * Nodes: Represent users, items, categories, etc.
    * Edges: Capture interactions such as likes, views, purchases, etc.
    * Example: In a movie recommendation system, nodes could represent users and movies, while edges represent viewing history and ratings.
2. Querying:
    * Graph Traversal Queries: Efficiently find relevant items by traversing the graph.
    * Example: To recommend books, a query might traverse from a user node to books they’ve rated highly, then to other users who rated those books highly, and finally to other books those users have rated highly.
3. Performance Optimization:
    * Indexing and Caching: Improve query performance by indexing frequently accessed nodes and caching results.
    * Example: Caching the results of common recommendation queries can reduce latency and improve user experience.

# How Graph Databases Support Better Recommendation Efficiency and Accuracy

Graph databases, such as [NebulaGraph](https://www.nebula-graph.io/), are particularly well-suited for handling the complex relationships and interconnected data involved in both ad serving and recommendation systems. Here’s how they support efficiency:

1. Complex Relationships and Connections:
    * Graph Structure: Graph databases store data in nodes (entities) and edges (relationships), making it easy to model and query complex relationships.
    * Example: Users, ads, and their interactions can be represented as nodes, with edges capturing relationships like views, clicks, and purchases. [LinkedIn](https://www.linkedin.com/) uses graph databases to model professional connections, endorsements, and job recommendations. By understanding the intricate relationships between users, companies, and skills, LinkedIn can provide highly relevant job suggestions and networking opportunities.
2. Real-Time Recommendations:
    * Fast Traversal: Graph databases excel at traversing relationships quickly, enabling real-time generation of recommendations based on user behavior and preferences.
    * Example: Quickly finding similar users or content based on shared attributes or interactions. [Pinterest](https://www.pinterest.com/) leverages graph databases to recommend pins based on user interactions and the connections between pins, boards, and users. This allows Pinterest to deliver personalized content in real-time, enhancing user engagement.
3. Personalization and Targeting:
    * Rich User Profiles: Graph databases can store and query detailed user profiles, including their connections, behaviors, and preferences, allowing for highly personalized ad targeting.
    * Example: Identifying users with similar browsing patterns and recommending ads that performed well with similar users. [eBay](https://www.ebay.com/) uses graph databases to improve its recommendation engine by analyzing user behavior, purchase history, and item relationships. This enables eBay to provide personalized product suggestions, increasing the likelihood of purchases.
4. Scalability:
    * Handling Large Datasets: Graph databases are designed to handle large volumes of interconnected data, making them suitable for high-traffic environments.
    * Example: Managing millions of users and ads, and their interactions, without performance degradation. [Facebook](https://www.facebook.com/)’s social graph, which models the relationships between billions of users, posts, and interactions, relies on graph databases to provide scalable and efficient recommendations. This ensures that users receive timely and relevant content, even as the platform grows.
5. Data Integration and Enrichment:
    * Combining Data Sources: Graph databases can integrate data from various sources, enriching user profiles and improving recommendation accuracy.
    * Example: Combining browsing history, social media activity, and purchase history to create a comprehensive user profile.
6. Feedback Loop and Continuous Learning:
    * Adaptive Algorithms: Graph databases facilitate the continuous updating of user profiles and recommendation algorithms based on real-time feedback.
    * Example: Adjusting recommendations based on recent user interactions to improve relevance.

# Conclusion
In summary, graph databases enhance the efficiency of ad serving and recommendation systems by providing a powerful way to model, store, and query complex relationships. They enable real-time personalization, effective targeting, and scalability, ultimately leading to better user engagement and optimized ad performance. As demonstrated by industry leaders like LinkedIn, Pinterest, eBay, and Facebook, leveraging graph databases can significantly improve user engagement and satisfaction.


This blog explores how recommendation systems work and how graph databases empower recommendation systems, providing real-life examples and architectural insights.

From Data to Insight: Leveraging Graph Databases for Superior Recommendations

This RideWallet is a leading force in the urban mobility payment industry worldwide. Integrated on the world’s renowned mobility service platform, the RideWallet has seamlessly facilitated millions of transactions, ensuring that passengers can effortlessly pay for their rides and other mobility services throughout the ride. **In addition to the ride payment, it supports a wide range of payment methods, including credit/debit cards, mobile wallets, loans and various local payment options tailored to regional needs.** In a short sentence, the RideWallet originated from the mobility service and evolved into a comprehensive payment solution.

# Rapid Expansion in Latin America Brings Financial Risk Management Challenges

In recent years, the RideWallet has been expanding rapidly, particularly in the Latin American market; for example, in Mexico alone, the company has issued over 5 million loans. However, with the boost of the business, it faces significant challenges in financial risk management. **The increasing geographical footprint has led to a surge in data volume, necessitating a system that can deliver faster response times, heightened accuracy, and unwavering stability.** The company encountered several critical challenges in its financial risk management operations:

- **Real-time Risk Control for Loan and New Card Issuance**: The consumer loan business, aimed at individual applicants, experiences high concurrency due to a large user base. To ensure a good user experience during the application process, the risk control decision engine cannot tolerate query delays exceeding one second. This places high demands on the graph technology's performance.

- **Mid-loan and Card Usage Risk Control**: Given the company's focus on the Latin American market, particularly Mexico, the relatively lenient overseas loan and credit card usage policies pose significant risks. Some payment institutions do not even verify CVV codes, leading to frequent misuse of loan and credit card limits. Traditional risk control rules struggle to accurately assess risks during the usage process.

- **Fraud Ring Detection**: For personal credit businesses, the risk control department is particularly concerned about fraud rings. Compared to individual loan application fraud, organized fraud poses a much greater threat. Therefore, accurately identifying high-risk communities is crucial to minimizing financial losses.

# NebulaGraph: Delivering Real-Time, High-Performance Risk Management Solutions

To address these complex challenges, NebulaGraph provided a comprehensive and efficient solution. NebulaGraph supports real-time graph construction, enabling full-graph risk control based on real-time application data. **With its high-performance data processing engine, the system can handle ultra-high concurrency, real-time bulk writes, and queries, ensuring stable operation even during peak times.**

**In terms of performance, NebulaGraph boasts exceptional computational capabilities, completing queries and calculations within the decision engine's tolerable time frame, ensuring the effectiveness of real-time rules.** The system employs advanced graph computation algorithms, combining real-time and asynchronous methods to extract relevant subgraphs from application data for asynchronous computation and then reintegrate them. This approach not only enhances processing efficiency but also ensures data accuracy and real-time performance.

Furthermore, NebulaGraph offers elastic scaling capabilities, allowing online scaling to ensure uninterrupted business operations. **The system can smoothly expand from 3 nodes to 12 nodes, and further scale online as needed to accommodate business growth, ensuring flexibility in response to increasing business demands.**

# The RideWallet's Use of NebulaGraph for Risk Control

The RideWallet leverages NebulaGraph in several key areas:

- **Real-time Risk Control During Loan Application**: Using the graph database, the system performs real-time queries to check if the applicant's IP address is a known blacklisted IP or if it connects to numerous devices. It also checks within three hops if the contact information provided by the user links to any blacklisted users.

- **Real-time Risk Control During Loan Usage**: Based on the fund flow graph, the system queries whether the user's loan funds are transferred to accounts that do not comply with the loan's intended use.

- **Risk Community Detection**: Using offline graph computations (WCC, Louvain), the system identifies communities closely linked to blacklisted nodes with high concentrations of black points. It extracts community features, such as the presence of multiple accounts with unidirectional transactions, a high proportion of midnight transactions, and several nodes with high PageRank values. These features are then fed into machine learning models for risk control decision-making.

![image](https://www-cdn.nebula-graph.io/blogs/RideWallet%20system%20architecture.png)
*System Architecture of the RideWallet*

# Achieving Significant Improvements with NebulaGraph's Advanced Capabilities

By adopting NebulaGraph's solution, RideWallet's credit and payment departments achieved significant improvements in multiple areas:

- **Business Value**: The enhanced risk control capabilities met the business expansion requirements, handling 3.3 billion edges and 630 million nodes. Through complex computations, the multi-stage risk control capabilities for application, credit approval, and loan usage based on graph technology were significantly improved, addressing data quality issues and enhancing overall risk control quality.

- **Technical Excellence**: NebulaGraph provided a stable, low-latency, high-throughput graph cluster capable of real-time updates, queries, and handling asynchronous computation tasks with a 4:9:1 ratio. Its distributed architecture and advanced graph computation technology ensured excellent performance in high-concurrency and complex computational environments, maintaining business continuity and stability.

- **Elastic Scalability**: NebulaGraph's elastic scaling capability allowed the system to smoothly expand from 3 nodes to 12 nodes, and further scale online as needed to accommodate business growth. This flexibility ensured the system could respond to business growth at any time, avoiding performance bottlenecks and service interruptions due to delayed scaling.

In summary, NebulaGraph's outstanding performance and powerful graph computation capabilities helped the customer achieve efficient real-time full-graph risk control, significantly enhancing both business and technical value.



This article introduce how a RideWallet leverages NebulaGraph for its Risk Management and the business results that NebulaGraph brings.