IS470

The SMU School of Information Systems (SIS) cordially invites you to see our students in action and find out how we integrate research into the curriculum here at SIS as part of the Research Experience for Undergraduates (RXU) via the  IS470 Guided Research in Information Systems module.

ABSTRACT Ciphertext-policy attribute-based encryption (CP-ABE) is a public key based one-to-many encryption that allows users to encrypt and decrypt data based on user attributes. A promising application of CP-ABE is flexible access control of encrypted data stored in semi-trusted servers. To reduce decryption overhead for users, we build an CP-ABE system with outsourced decryption. In such a system, a user provides a semi-trusted server with a transformation key that allows the server to translate any ciphertext into a simple ciphertext, and the user can decrypt the simple ciphertext with little overhead. In this project, we design and implement a secure group chat system using CP-ABE, which includes three generic parts: Key Generation Server, Messaging Server, and Client. The Key Generation Server is offline, and is used to generate keys and manage users’ attributes. The sender  encrypts a chat message and sends the ciphertext to the Messaging Server. The Message Server is responsible for storing and performing outsourced decryption of encrypted chat messages. This project extends an IS 470 project in the last term with two major differences: the first is that the system runs on Android mobile platforms rather than on Windows computers; the second difference is that the system makes the CP-ABE chat as a plug-in while the previous system embeds the CP-ABE code into an insecure chat system.

ABSTRACT We survey the notion of provably secure searchable encryption (SE) in this project. In particular, we provide a comprehensive overview of two main SE techniques: searchable symmetric encryption (SSE) and public key encryption with keyword search (PEKS). We focus on two pioneering works by Song et al. and Boneh et al. and categorize other works. Different SE schemes are compared in terms of their security, efficiency and functionality. We also show the connections between many other approaches to SE and identify open research problems.

ABSTRACT This study seeks to examine how Machine Learning algorithms can be employed to automate the process of reviewing and providing feedback to programming assignments submitted by students. The presentation will demonstrate how student lab test submissions can be pre-processed and analysed using clustering algorithms, and the performance measures used to evaluate the effectiveness of these techniques. It also explores the potential uses of these findings in improving the efficiency and effectiveness of the feedback process, as well as some points for further research.

ABSTRACT Technology plays a critical role in managing nowadays large scale and automated trading activities in stock exchanges. The development of information systems facilitates the instantaneous information exchange among various parties within the finance domain. As a result, participants across different time zones and physical boundaries are involved in trading activities in one integrated exchange platform. As the trading decisions grow more and more complex, stock exchanges have to constantly upgrade their systems to meet the emerging needs of clients. This presentation examines how information systems and architecture in stock exchange enable the fast and automated trading processes. It first introduces the typical trading cycles and business processes involved within stock exchange. It discusses the different information systems and their functions in those processes. It then analyses the architecture of stock exchange using one case study. Also, it concludes the good architecture and practices in exchange development and discusses the future trends of e-trading. Lastly, it shows an exchange prototype and gives a demo.

ABSTRACT We want to automatically determine relative code quality by looking at the maintainability of parts of a system and make better decisions on alternative designs. In order to do this, we developed a tool to generate the dependency graph model and compare change commits to collect data on various projects to see if it is possible to estimate the size of the ripple set from the size of the dependency set, where the dependency set is derived from non-transitive directed syntactic dependencies of the seed modules. We ran the analysis on open-source Java projects, such as log4j. Our results show that dependency set is a better estimate than the fully transitive model and may be used to estimate the relative modifiability between competing designs, given a requirement change. Historically, this has not been possible as change impact analysis to determine modifiability is done after changes have been made, and not before. Our tool is ideal because we are able to determine the effort required to make a given change before making any modifications to the codebase, and hence, it is better than previous models.