Cheng-Zen Yang

An iterative approach for web catalog integration with support vector machines

Web catalog integration is an emerging problem in current digital content management. Past studies show that more improvement on integration accuracy can be achieved with advanced classifiers. Because Support Vector Machine (SVM) has shown its supremeness in recent research, we propose an iterative SVM-based approach (SVM-IA) to improve the integration performance. We have conducted experiments of real-world catalog integration to evaluate the performance of SVM-IA and cross-training SVM. The results show that SVM-IA has prominent accuracy performance, and the performance is more stable.

On hierarchical web catalog integration with conceptual relationships in thesaurus

Web catalog integration is an interesting problem in current digital content management. Past studies have shown that using a flattened structure with auxiliary information extracted from the source catalog can improve the integration results. However, the nature of a flattened structure ignores the hierarchical relationships, and thus the performance improvement of catalog integration may be reduced. In this paper, we propose an enhanced hierarchical catalog integration (EHCI) approach with conceptual thesauri extracted from the source catalog. The results show that our enhanced hierarchical integration approach effectively boosts the accuracy of hierarchical catalog integration.

Information retrieval on bug locations by learning co-located bug report clusters

Bug locating usually involves intensive search activities and incurs unpredictable cost of labor and time. An issue of information retrieval on bug locations is particularly addressed to facilitate identifying bugs from software code. In this paper, a novel bug retrieval approach with co-location shrinkage (CS) is proposed. The proposed approach has been implemented in open-source software projects collected from real-world repositories, and consistently improves the retrieval accuracy of a state-of-the-art Support Vector Machine (SVM) model.

CS-SCTP: A collaborative approach for secure SCTP over wireless networks

Stream control transmission protocol (SCTP) is recently proposed as a transport layer protocol with many attractive features. However, its native design does not consider the wireless operations and the security demands. Although past studies were conducted to improve either its wireless performance or its security functionality, we cannot And any research considering both issues. In this paper, we proposed a collaborative approach (CS-SCTP) for secure SCTP communication in a wireless environment. CS-SCTP employs two mechanisms, partial authentication and two- level key encryption, to preserve the performance advantages in Collaborative SCTP (C-SCTP) and the security enhancement in Secure SCTP (S-SCTP). In the simulations, the performance results show that CS-SCTP outperforms S-SCTP and is comparable to C-SCTP.

RDF/XTM ontology construction based on a topic maps-driven framework

Ontology construction plays an important role in many semantically working environments, such as knowledge management systems and digital libraries. The efficiency, quality, and comprehensiveness of the construction process highly depends on the employed development framework. In this paper, we propose a Topic Maps-driven framework, called XRVAT, to alleviate the cumbersome construction procedure and generate both RDF/XTM ontologies with an effective XTM-to-RDF translation kernel. This paper briefly presents its design and an operational example.

Design of a QoS gateway with real-time packet compression

Sufficient network bandwidth and smooth communication tunnels determine the success of Internet applications. However, the growth of bandwidth requirements is much faster than the growth of physical network bandwidth. To manage limited network bandwidth, the QoS service is an effective approach. Nevertheless, the performance of previous QoS services is limited by the physical network bandwidth. In this study, we propose a QoS gateway with real-time packet compression. The proposed QoS gateway provides the packet transmission rate higher than the physical network bandwidth. To evaluate the proposed QoS gateway, we have implemented a prototype on Intel IXP425 platform. We adopted the LZSS compression algorithm and provided several LZSS implementations for performing differential services. The experimental result shows that the transmission rate increases 15%.

Design and Implementation of a Mobile SSH Protocol

SSH (secure shell) is a very popular secure communication application. Users can remotely login to servers through SSH channels with security protection. However, since the SSH protocol lacks mobility management support, an SSH session can be easily broken while the wireless device roams around. In this paper, we propose a mobile SSH protocol which employs an application-layer handover mechanism. We have implemented the proposed SSH protocol in OpenSSH and Putty. With the proposed protocol, the mobile users can switch from one wireless network to another without terminating their SSH sessions. This paper presents the design of the proposed mobile SSH protocol and the preliminary experimental results

A novel Java RMI middleware design for active networks

Java RMI (remote method invocation) facilitates the complicated procedure of developing large-scale distributed applications. However, this paradigm suffers from a centralized bottleneck problem when many clients send requests to a server simultaneously. We propose a novel middleware design called ActiveRMI to enhance RMI performance by employing active networks infrastructure with a code caching scheme. With ActiveRMI, remote services can be dynamically cached and built up at intermediate active routers to process further client requests as soon. Therefore, active RMI achieves three improvements: (1) the RMI performance is improved; (2) system availability is increased and (3) system load and network traffic is reduced. As the preliminary experiments show, programmers can use the code generation tool ARMIGen to design ActiveRMI applications effectively and efficiently. Furthermore, the user response time of RMI applications can be highly improved. The promising results conduce to the future development of next generation middleware on active networks.

Improving hierarchical taxonomy integration with semantic feature expansion on category-specific terms

In recent years, the taxonomy integration problem has obtained much attention in many research studies. Many sorts of implicit information embedded in the source taxonomy are explored to improve the integration performance. However, the semantic information embedded in the source taxonomy has not been discussed in the past research. In this paper, an enhanced integration approach called SFE (Semantic Feature Expansion) is proposed to exploit the semantic information of the category-specific terms. From our experiments on two hierarchical Web taxonomies, the results are positive to show that the integration performance can be further improved with the SFE scheme.

Learning to integrate web catalogs with conceptual relationships in hierarchical thesaurus

Web catalog integration has been addressed as an important issue in current digital content management. Past studies have shown that exploiting a flattened structure with auxiliary information extracted from the source catalog can improve the integration results. Although earlier studies have also shown that exploiting a hierarchical structure in classification may bring better advantages, the effectiveness has not been testified in catalog integration. In this paper, we propose an enhanced catalog integration (ECI) approach to extract the conceptual relationships from the hierarchical Web thesaurus and further improve the accuracy of Web catalog integration. We have conducted experiments of real-world catalog integration with both a flattened structure and a hierarchical structure in the destination catalog. The results show that our ECI scheme effectively boosts the integration accuracy of both the flattened scheme and the hierarchical scheme with the advanced Support Vector Machine (SVM) classifiers.