1 April 2024 in Proceedings of the VLDB Endowment
Discovering tables from poorly maintained data lakes is a significant challenge in data management. Two key tasks are identifying joinable and unionable tables, crucial for data integration, analysis, and machine learning. However, there's a lack of a comprehensive benchmark for evaluating existing methods. To address this, we introduce LakeBench, a large-scale table discovery benchmark. It evaluates effectiveness, efficiency, and scalability of table join & union search methods. With over 16 million real tables, LakeBench is 1,600X larger than existing datasets and 100X larger in storage size. It includes synthesized and real queries with ground truth, totaling more than 10 thousand queries - 10X more than used in any existing evaluation. We spent over 7,500 human hours labeling these queries and constructing diverse query categories for thorough evaluation. Our benchmark thoroughly evaluates state-of-the-art table discovery methods, providing insights into their performance and highlighting research opportunities.
1 January 2019 in International Journal of Information and Communication Technology
The problems and efficient in existing intelligent transportation system (ITS) models are studied and the significance of the model and architecture of ITS under internet of things (IoS) are also discussed. By the research status at home and abroad about the relationship between IoS and ITS, the necessity of IoS technology introduce to ITS is explained first. Then, based on logical structure and physical model of ITS, the ITS structure model under IoS environment is established. Furthermore, the system architecture on a prefect and comprehensive ITS information system acquisition system is formed. The design principles of system demand, overall design, function modules design, database design are described in detail. The parts of key modules in system are introduce and tested. The tests results show the acquisition system can effectively monitor real-time of vehicle speed, real-time traffic of road vehicles and obtain effective information of vehicle and road environment. It can also send the information acquired via internet or GPRS network to data processing centre, for future process and intelligent decision of ITS.
1 July 2026 in Review of Scientific Instruments
Accurate characterization of the three-dimensional (3D) spatial radiation characteristics of microwave fields, especially the out-of-plane intensity distribution, remains a critical challenge in advanced microwave technology and diamond nitrogen-vacancy (NV) center-based quantum sensing. In this work, a high-precision three-dimensional microwave magnetic field measurement system based on a fiber-integrated NV center scanning probe is developed. The system enables high-spatial-resolution characterization of electromagnetic fields on the surface of miniaturized microwave antennas. The core of the system is a mechanically polished tapered optical fiber with a 26.5° taper angle, integrated with a diamond nanopillar containing ensemble NV centers at the fiber tip. This tapered structure conforms more closely to the diamond nanopillars, thereby enabling high-sensitivity electromagnetic imaging on the chip surface. We perform microwave magnetic field measurements on a standard Ω-ring radiating structure: in-plane measurements at 100 and 500 μm from the ring edge verify excellent field uniformity in the central region of the Ω-ring radiating structure, while out-of-plane characterization from 0 to 500 μm above the ring center identifies an effective radiation distance of 500 μm for the structure. This system realizes non-destructive, high-precision on-chip electromagnetic field characterization and can be extended to diverse quantum magnetometry applications requiring 3D measurement capability.
1 September 2024 in Review of Scientific Instruments
A diamond nitrogen-vacancy (NV) ensemble has been developed as a vector magnetometry platform for sensing external time-varying magnetic fields. However, due to the complexity of manipulating electron spins along different directions, a current vector NV magnetometer often needs a large amount of supporting equipment, preventing its applications in a compact circumstance. Here, we develop a hardware-level protocol to realize a multi-axis NV magnetometer using only a single channel of microwave generation and signal detection resources. This mechanism is to monitor each resonance serialized in a sequence and measure the electron-spin frequency shifts concurrently in real time. The functionality is realized by a home-made control system with an on-chip direct digital synthesis generator and signal processor. We finally achieve a vector sensitivity of around 14 nT/Hz on four different axes at the same time. We also analyze the phase delay of the sensing signal between different axes induced by the protocol. This protocol is compatible with other schemes to further improve the performance, such as hyperfine driving, balanced detection, and high-efficiency photon collection methods.