Wenqi Jiang

A student scholarship for the development of linking algorithms and graph linkers.

• Analyzed ontologies from different data sources, like DBpedia, Wikidata;
• Specialized 7 linking discovery algorithms, including string-based similarity and geometry relationships;
• Implemented a Sparql-based linking application by using Apache Jena, and evaluated it with the OAEI-2021 datasets;
• Written the master’s thesis.

Keywords: Ontology Graph Linker RDF Text Similarity DE-9IM WKT SPARQL Apache Jena OAEI

Shenzhen Togic Software Technology Co. was established in 2010. The company focuses on home Internet TV clients, with its Togic (similar to Netflix) and WEBOX (similar to Apple TV) delivering a simple, intuitive, and lightweight experience for family entertainment, quickly becoming the most popular app on TV systems. There are currently near 500,000 daily users on this platform, and over 9 million customers worldwide used these products.

Offline Data Warehouse

• Maintained the Hadoop cluster containing over 200 TB of data, including user and system logs from 3 different sources;
• Upgraded big data components, including Hadoop, Spark, Kafka, etc., and monitored them on CDH;
• Designed and implemented a four-layer data warehouse based on the STAR model by using Hive, which had near 150 tables and received over 300 GB per day;
• Rewrote the ETL pipeline, which saved over 80% of storage and fixed the small files’ problem on HDFS.

Keywords: CDH Kafka Hadoop HIVE Spark

Data Statistics System

• Restructured the statistics system based on the new offline data warehouse;
• Rewrote the code by using Spark DSL, making it more readable and maintainable;
• Optimized the system’s performance, reduced the process time by 25% compared to the previous version;
• Implemented more than 30 new statistical demands and fixed bugs.

Keywords: Sqoop Hive Spark Scala Elasticsearch MySQL MongoDB Python

Log Collection Service

• Designed and implemented the log collection pipeline using Nginx and Netty, improving performance, stability, and scalability;
• Processed all logs into the Kafka cluster, which has 5 brokers, 3 replications, and 7 days of redundancy, improving robustness.

Keywords: Nginx Netty Java Kafka HDFS

Beijing Webstudio Information Technology Co. was founded in 2007 and focuses on quality-based and innovative development, providing data analysis and visualization, as well as official website marketing solutions which span a wide range of business disciplines and application situations.

• Implemented backend APIs of big data visualization dashboards and general reports for the BI applications;
• Developed more than 25 data operators, such as summation, date conversion, string processing etc.;
• Implemented the one-click function of syncing all user information;
• Expanded 16 data sources, including MySQL, SQL Server, MongoDB etc.;
• Maintained the big data cluster;
• Fixed bugs and developed new features for the previous company’s projects.

Keywords: Dubbo Spring Boot MyBatis Hibernate Java Nginx MySQL Redis Docker FastDFS CDH Hadoop HBase Impala Spark ML