Guangming Xing

Support vector machines for collaborative filtering

Support Vector Machines (SVMs) have successfully shown efficiencies in many areas such as text categorization. Although recommendation systems share many similarities with text categorization, the performance of SVMs in recommendation systems is not acceptable due to the sparsity of the user-item matrix. In this paper, we propose a heuristic method to improve the predictive accuracy of SVMs by repeatedly correcting the missing values in the user-item matrix. The performance comparison to other algorithms has been conducted. The experimental studies show that the accurate rates of our heuristic method are the highest.

Teaching software engineering using open source software

This paper describes our experience of using open source software systems in teaching a graduate level software engineering course. The motivation of this course, the course structure, the assessment, and the outcomes are discussed. The comparative results using different approaches are also presented.

X2R: a system for managing XML documents and key constraints using RDBMS

We describe X2R, an XML document management system which supports efficient storage, retrieval and key constraints for XML documents. The system is based on a mapping algorithm that translates a DTD to a relational schema. Based on the mapping, node groups are range indexed and shredded into the database. It has been shown that queries are executed several times faster than using the methods in literature, and space usage is significantly reduced. We have also added key support to the system by propagating key constraints for XML documents to keys in a relational schema. The system was designed for efficient access of the vast amount of environmental data, which are widely available in XML format.

An optimization method for selecting parameters in support vector machines

It has been shown that the cost parameters and kernel parameters are critical in the performance of support vector machines (SVMs). A standard parameter selection method compares parameters among a discrete set of values, called the candidate set, and picks the one which has the best classification accuracy. As a result, the choice of parameters strongly depends on the pre-defined candidate set. In this paper, we formulate the selection of the cost parameter and kernel parameter as a two-level optimization problem, in which the values of parameters vary continuously and thus optimization techniques can be applied to select ideal parameters. Due to the non-smoothness of the objective function in our model, a genetic algorithm has been presented. Numerical results show that the two-level approach can significantly improve the performance of SVM classifier in terms of classification accuracy.

Efficient schema extraction from a large collection of XML documents

XML is becoming the standard format for data exchange on the Internet. In this paper, we present a system that is effective in extracting schema information from a large collection of XML documents. Based on Xtract, we propose using the cost of an NFA simulation to compute the Minimum Length Description. We also studied using frequencies of the sample inputs to improve the effectiveness of the schema extraction. Experimental studies were conducted on synthesized XML data sets, suggesting the efficiency and effectiveness of our approach as a solution for schema inference.

Index and storage design in native XML databases

Using a native XML DBMS to efficiently store and retrieve XML data has been well studied in recent years. Various native XML DBMSs were developed, and they support most of the features of XQuery. The core topics of these studies are indexing methods for XML. Most existing native XML DBMSs use range based indexing methods, resulting in significant extra space and system degradation.In this paper, we present a method that maps XML data into relational tables. Our method is still range based, which means it can process queries with regular path expression efficiently. But with the introduction of node grouping, the extra space needed for the indexing is reduced significantly. With fewer tables created, we will show that query processing can also be significantly expedited.

Approximate matching of XML documents with schemata using tree alignment

Studying structural similarity between XML documents are important for many tasks like XML data classification and XML document management. In this paper, we formally introduce the edit distance between trees and tree grammars using tree alignments. A sketch of the proof for the correctness of the algorithm is presented followed by the analysis of the efficiency of the algorithm. Experiments are conducted to show the time efficiency and validness of the distance in the context of XML document clustering.

Organizational search in email systems

With the storage space for emails increasing everyday, email systems are serving as a data repository for many users. It is important to offer intuitive and flexible search utilities to provide better access to the information in the emails, especially in an enterprise setting. In this paper, we present TESO, a tool for Email Searching using Organizational information. TESO seeks to improve the relevancy of email search by integrating data from email servers, and organizational information from directory services and other resources. TESO has been developed as an add-on for mozilla thunderbird. We have experimented using SQLLite as well as XML data to store the organizational information. Our studies also provide a solid test bed for existing algorithms for information integration and keyword search in relational databases and XML files.

An excel based Semantic email system

Semantic Email is a type of Semantic Web application, which deals with the understanding of emails received, and performing corresponding actions according to the schema specified in the Semantic Email system.In this paper, the Semantic Email is implemented with a restriction of using excel sheets as a mode of communication. The current Semantic Email project could be reengineered to support heterogeneous semantic actions based on the action plan. The Semantic Email project can be enhanced providing a web interface, apart from the email system currently used for communication. The clients can directly use the web page, corresponding to the schema rather than sending an email.

Improving access control for browsers using dynamic scoping

Web is playing a more and more important role in our daily life. Large volume of personal and business communications are taking place via Web everyday. Due to its importance and its vulnerabilities from the early design, the Web has become a preferred target of attacks. After discussing the causes and effects of vulnerabilities, we then present that the current access control system of the web is the root cause for these problems. As a part of the efforts to develop a fine-grained access control system in a Web application, we focus on the dynamic scoping for browser based access control. Instead of using static scope to isolate the client code from the trusted web application, dynamic scoping is utilized to identify different patties in a Web page. Such an improvement will allow the client and trusted web application share common libraries, while still get executed at different trust levels.