UROP Proceedings 2022-23

School of Engineering Department of Computer Science and Engineering 124 Knowledge Discovery over Database Supervisor: WONG, Raymond Chi Wing / CSE Student: WANG, Shaoqi / COMP Course: UROP1100, Spring The objective of this report is to detail the work I completed as part of the Knowledge Discovery Over Database project, which involved studying two courses: COMP3711 - Design and Analysis of Algorithms and COMP2012 - Object-Oriented Programming and Data Structures. Throughout the project, I accumulated knowledge of commonly used algorithms such as divide-and-conquer, greedy algorithms, dynamic programming, and graph. I also learned object-oriented programming techniques, including inheritance, polymorphism, and templates, to enhance the code's readability, reusability, and scalability. In this report, I will illustrate the important knowledge I learned in these two courses and how they are possibly related to time series analysis. It also concludes with a summary of the insights I gained from the project and my perspectives on the future of this project. Knowledge Discovery over Database Supervisor: WONG, Raymond Chi Wing / CSE Student: XU, Zian / DSCT Course: UROP1100, Fall The recommendation system plays an important role in our daily life, especially when we search randomly online, waiting for interesting items to appear. Traditional recommendation system uses the information of users’ past interests. However, a study shows that users’ interest sometimes changes a lot from time to time, while it tends to be consistent in a short term. In view of this, the session-based recommendation system has caught people’s attention, which makes predictions with information in a small session. However, most of the session-based models ignored the timestamp of user actions that may contain meaningful information. In this report, I will show two methods to add time interval information to a GNN model and some analysis of the result. Research on Mining Course Structure Supervisor: WONG, Raymond Chi Wing / CSE Student: GU, Yuyang / SSCI Course: UROP1100, Summer With the development of the Internet, people can access an increasing amount of information in their daily lives. To address this phenomenon of "information explosion," recommendation systems have been widely applied in various areas to assist users in filtering information. In recent years, numerous algorithms have been developed to identify the most suitable choice with the highest utility from a given database. However, these algorithms primarily focus on choices with independent attributes, and research on data with dependent attributes is scarce. Inspired by this, I have dedicated the past weeks to studying this problem. In this report, I will provide a rigorous definition of this particular problem and present one approach I have explored to solve it.