Hong, Liangjie > Research

Current Research Topics:

Birds of a Feather Project

Jan, 2008-Present

Directors:
H. Wagner Professor Roger Nagal
Assistant Professor Brian D. Davison

The goal of the birds of a feather project is, to develop a white paper that clarifies the enhanced value propositions offered by using social networks and virtual reality as environments for business development and globalization. To provide an understanding of the current and potential benefits, as well as the constraints imposed by cultural, technological, and business trends that must be dealt with in order to achieve sustainable strategic benefits.

The industrial users are bottom line focused; they seek not new methodologies, but enhanced and sustained value creation. The paper will include mini case examples of what leading companies are now doing. In addition an integrated view of future business applications will be extrapolated from emerging trends, interviews with business leaders, technology experts, and related sources.

I am working in the project and participating the environment evaluation part and also moving onto the impact of the computer science on social network services part.

Understanding Discussion Boards by Speech Acts

Jan, 2008-Present

This is the project for course Web Mining instructed by Assistant Professor Brian D. Davison

Discussion boards are very popular in virtual internet communities and have many applications in areas such as customer support, community development, interactive reporting and education. Since discussion boards contain numerous of threads related to problem solving, one intuitive research idea based on discussion boards is to extract knowledge from discussion boards especially questions and their corresponding solutions. This work can be called as building a Question-Answering (QA) system from discussion boards.

In traditional QA Systems, emerged from 1960s, questions are provided by users, normally one-sentence questions. Most researchers focused on how to mining answers from large repositories. Little research focused on how to find or detect questions, which is totally unnecessary for those systems. However, things become different for discussion boards. In discussion boards, not every thread is a question thread. Besides, the form of questions in discussion boards varies significantly from traditional questions in QA Systems while they are in totally free form like conversations rather than one-sentence questions. Furthermore, questions may have several sentences or even multiple paragraphs, often companied by examples and scenarios. If we cannot identify whether a thread is about one certain question, answer mining becomes meaningless.

One future benefit from questions detection is that if question-solution pairs can be extracted from discussion boards, it is easy to build wiki-like FAQ or support website automatically from the pairs.

Previous Selected Research Activities:

Port Virtual Routing Ring (VRR) into NS-2

As the class project of Topics in Communication Networks course, I ported all the source code of Virtual Routing Ring protocol (VRR), a novel routing protocol for Ad-Hoc network developed by Microsoft Research, into NS-2 environment while the original source code written in C and C# is for Windows environment.

Pattern Recognition Practice

As the class project of Pattern Recognition course, I used several Naïve-Bayes classifier to do character recognition on different quality of character images. This project finally gave a deep understanding of pattern recognition theories and techniques.

myXML Studio

This project is to provide a development environment like Eclipse for all kinds of XML files with plug-in architecture for future extensions.

My job in the project was to make the design of the basic architecture of the software and implement the Eclipse-like plug-in architecture.

Campus Search Engine

The goal of this project is to integrate automatically information on campus websites.

Served as the Director of the project, I designed and implemented a storage solution for thousands of web pages rather than existing databases. The search engine has several independent components that could be deployed in multiple computers for distributed computing. The project, funded by the university, was completed by a four-student team.

[Project Report] Warning! This document is in Chinese.

Simple Compiler & its Generator

The aim of the project is to build tools like Bison, Lex and Yacc for generating scanners and analyzers for a programming language’s compiler.

I built the compiler generator and actually generated the scanner and analyzer using the generator I implemented.

Student Community Online

The aim of the project is to establish a new version of community, including BBS, weblog, information publishing system and learning system for students and faculty in the university.

Served as the Director of the project, I organized the whole team and created the basic design. The project is funded by Student Network Center (BUCT) and students conducted all the operation and development of the whole project. The project team consisted of more than 10 students.