School of Engineering Department of Computer Science and Engineering 130 Large-Scale Spatiotemporal Data Analytics and Learning Supervisor: ZHOU, Xiaofang / CSE Student: LI, Yunqi / COSC Course: UROP1100, Fall UROP2100, Spring This report is a short brief for our paper about efficiency and privacy protection-oriented graph query framework FRESH. We study the efficiency and privacy protection in graph query over large databases. We will start from the graph anonymization process, then come to three classical graph problems subgraph matching, shortest distance query and triangle counting to tap into the space where these algorithms can be improved. Then we come to the mathematical model to describe the cost for the framework and draw a conclusion. Finally we present the experimental results in appendix. Large-Scale Spatiotemporal Data Analytics and Learning Supervisor: ZHOU, Xiaofang / CSE Student: MURUGU, Hansika / COMP Course: UROP1100, Fall Spatiotemporal data describes moving entities which are captured across geographical spaces and quantifiable time. The applications of spatiotemporal data analytics are vast such as enabling intelligent transport systems, location-based services, resource tracking and scheduling, emergency responses, urban planning, IoT, and smart cities among others. This project aims to conduct a methodical data analysis in order to identify bottle necks in the procedure and develop meaningful insights. The preliminary research efforts is focusing on analysing and visualizing the data collected from the cargo logistics system of the Hong Kong International Airport. Large-Scale Spatiotemporal Data Analytics and Learning Supervisor: ZHOU, Xiaofang / CSE Student: SU, Chen-yi / DSCT Course: UROP1100, Spring This study aims to evaluate the performance of various pathfinding algorithms, such as Dijkstra's, bidirectional Dijkstra, and A* algorithm, in different graph scenarios, considering factors like graph complexity, distance between points, and graph size. Utilizing OpenStreetMap data, the project analyzes the performance of both custom and NetworkX library implementations of these algorithms for short and long distances within a 6 km radius. The results are visualized through a web application, enabling a comprehensive understanding of algorithm behavior and performance under varying graph conditions. The findings provide valuable insights into the applicability and efficiency of these pathfinding algorithms in realworld navigation systems, ultimately guiding the selection of the most suitable algorithms for specific use cases.
RkJQdWJsaXNoZXIy NDk5Njg=