Yuhai Zhao

A Self-healing composite Web service model

Composite Web services are often long-running, loosely coupled and cross-organizational applications. They always run in a highly dynamic environment. For the applications and environment, advanced transaction support is required to ensure the quality of reliable execution. Towards composite service adaptive mechanism unavailable for lacking transaction support, this paper proposes a self-healing model for Web service reliable execution, which is an integration of flexible compensation service in selection and reselecting in execution. In order to make the composite service healing itself as quickly as possible and minimize the number of reselections, away of mining cascading scope of replacement in advance by considering full multi-relation among transaction Web services is proposed in this paper. Further more, a new comprehensive, objective QoS-driven service replacement model with compensation support is presented, and the self-healing algorithm is proposed. Experiments show that the model guarantees business process reliability.

Mining time-shifting co-regulation patterns from gene expression data

Previous work for finding patterns only focuses on grouping objects under the same subset of dimensions. Thus, an important bio-interesting pattern, i.e. time-shifting, will be ignored during the analysis of time series gene expression data. In this paper, we propose a new definition of coherent cluster for time series gene expression data called ts-cluster. The proposed model allows (1) the expression profiles of genes in a cluster to be coherent on different subsets of dimensions, i.e. these genes follow a certain time-shifting relationship, and (2) relative expression magnitude is taken into consideration instead of absolute one, which can tolerate the negative impact induced by noise. This work is missed by previous research, which facilitates the study of regulatory relationships between genes. A novel algorithm is also presented and implemented to mine all the significant ts-clusters. Results experimented on both synthetic and real datasets show the ts-cluster algorithm is able to efficiently detect a significant amount of clusters missed by previous model, and these clusters are potentially of high biological significance.

Finding Top-k Covering Irreducible Contrast Sequence Rules for Disease Diagnosis

Diagnostic genes are usually used to distinguish different disease phenotypes. Most existing methods for diagnostic genes finding are based on either the individual or combinatorial discriminative power of gene(s). However, they both ignore the common expression trends among genes. In this paper, we devise a novel sequence rule, namely, top-k irreducible covering contrast sequence rules (TopkIRs for short), which helps to build a sample classifier of high accuracy. Furthermore, we propose an algorithm called MineTopkIRs to efficiently discover TopkIRs. Extensive experiments conducted on synthetic and real datasets show that MineTopkIRs is significantly faster than the previous methods and is of a higher classification accuracy. Additionally, many diagnostic genes discovered provide a new insight into disease diagnosis.

An Active Service Reselection Triggering Mechanism

Actively identifying service faults and actively trigger service reselection are important management methods for efficiently promoting system reliability. Composite Web services, however, are often long-running, loosely coupled and cross-organizational applications. They always run in a highly dynamic and changing environment(The Web) which imposes many uncertainties, such as server unavailable or network interruption or temporarily interrupt and so on. Under these uncontrollable circumstances, it is impractical to monitor the changes in Quality of Service parameters for each and every service in order to timely trigger service reselection, due to high computational costs associated with the process. In order to overcome the above problem, this paper proposes an efficient reselection mechanism by mining early patterns in advance. The system will trigger service reselection once potential execution failure is detected by matching these early patterns with the current situation. The process is real time and low-cost, and as such, the proposed mechanism will improve system reliability.

A novel approach to revealing positive and negative co-regulated genes

As explored by biologists, there is a real and emerging need to identify co-regulated gene clusters, which include both positive and negative regulated gene clusters. However, the existing pattern-based and tendency-based clustering approaches are only designed for finding positive regulated gene clusters. In this paper, a new subspace clustering model called g-Cluster is proposed for gene expression data. The proposed model has the following advantages: 1) find both positive and negative co-regulated genes in a shot, 2) get away from the restriction of magnitude transformation relationship among co-regulated genes, and 3) guarantee quality of clusters and significance of regulations using a novel similarity measurement gCode and a user-specified regulation threshold δ, respectively. No previous work measures up to the task which has been set. Moreover, MDL technique is introduced to avoid insignificant g-Clusters generated. A tree structure, namely GS-tree, is also designed, and two algorithms combined with efficient pruning and optimization strategies to identify all qualified g-Clusters. Extensive experiments are conducted on real and synthetic datasets. The experimental results show that 1) the algorithm is able to find an amount of co-regulated gene clusters missed by previous models, which are potentially of high biological significance, and 2) the algorithms are effective and efficient, and outperform the existing approaches.

An Efficient and Effective Overlapping Communities Discovery Based on Agglomerative Graph

Community discovery is a popular way to solve the personal service recommendation problem and has recently attracted more and more attentions of the researchers. The communities are often practically overlapping with each other, thus more and more research focus on the problem of overlapping communities detection. A common drawback of the existing algorithms to this problem is the low efficiency when dealing the large scale network. In this paper, we propose a graph compression based overlapping communities discovery algorithm, which greatly enhances the power of handling large networks even using a single computer. First, a graph compression based social network model, namely agglomerative graph, is introduced, which is a lossless compression to the original network. Then, inspired by the idea of iteration based on the selected seeds, the algorithm expands the selected seeds to the communities by optimizing the proposed community fitness function iteratively. Finally, it merges the communities of high similarity with each other to get the final results. Since the network is lossless compressed, and massive redundant computations are avoided, the results can be exactly obtained in an efficient and effective way. The experiments based on both real and synthetic datasets demonstrate efficiency and effectiveness of the proposal method in detecting overlapping communities over large scale networks.

GDC: An efficient tag recommendation algorithm

Diversifying the query results is an effective approach to obtain an users underlying query information. Tag as an important recommendation information has drawn more and more attentions, which is also extensively exploited in SN(Social Network). The results generated by the current algorithms often ignore diversity in the recommended results, which seriously affecting the users experience. In this paper, we address the problem of diversified coverage based tag recommendation. To our best knowledge, it is first proposed in the context of tag recommendation. Since the NP-hardness of the problem, a greedy based diversified coverage tag recommendation algorithm, namely GDC, is proposed. First, it constructs a semantic similarity graph based on the local and global tag co-occurrence matrices, which improves the recommendation accuracy by incorporating both the users interests and the popularity of tags. Further, GDC recursively searches the MIDS (Minimum Independent Dominating Set) in an efficient manner together with an effective pruning rule. The experiments conducted on the real datasets of MovieLens and Last.fm show that the proposed GDC improve the diversity significantly.

Improving ELM-Based Service Quality Prediction by Concise Feature Extraction

Web services often run on highly dynamic and changing environments, which generate huge volumes of data. Thus, it is impractical to monitor the change of every QoS parameter for the timely trigger precaution due to high computational costs associated with the process. To address the problem, this paper proposes an active service quality prediction method based on extreme learning machine. First, we extract web service trace logs and QoS information from the service log and convert them into feature vectors. Second, by the proposed EC rules, we are enabled to trigger the precaution of QoS as soon as possible with high confidence. An efficient prefix tree based mining algorithm together with some effective pruning rules is developed to mine such rules. Finally, we study how to extract a set of diversified features as the representative of all mined results. The problem is proved to be NP-hard. A greedy algorithm is presented to approximate the optimal solution. Experimental results show that ELM trained by the selected feature subsets can efficiently improve the reliability and the earliness of service quality prediction.

Dynamic Self-Healing Mechanism for Transactional Business Process

It is clear that transactional behavior consistency is a prerequisite and basis for construction of a reliable services-based business application. However, in previous works, maintaining transactional consistency during exception handling was ignored. Maintaining transactional consistency requires functionality for rolling back some operations and revoking uploaded data. Replacing only the failed service will eventually lead to overall business application failure. In this study, we take fully into account the behavioral consistency of transactional services and propose two effective self-healing mechanisms for service-based applications. If a service enters into potential failure condition, a rescheduling mechanism is triggered to maintain consistent transactional behavior and to ensure reliable execution; if a service fails during execution, the compensation operation is triggered and the system will take action to ensure transactional behavior consistency. Meanwhile, cost-benefit analysis with compensation support is proposed to minimize the dynamic reselection cost. Finally, the experimental analysis shows that the proposed strategies can effectively guarantee the reliability of Web-based applications system.

Identifying top-k Vital Patterns from multi-class medical data

With the development of modern science, the goal of medical research is not limit to explore a type of disease but more accurate multi-subtypes of this disease. For example breast cancer can be divided into three different subtypes: BRCA1, BRCA2 and Sporadic. Previous work only focuses on distinguishing several pairs of tumors. However, the simultaneous distinguish across multiple disease types has not been well studied yet, which is important for medical researcher. In this paper, we define VP (an acronym for “Vital Pattern”) and PP (an acronym for “Protect Pattern”) by a statistical metric, and propose a new algorithm to make use of the property discovery VP and PP from multiple disease types. The algorithm can generate some useful rules for medical researchers. The results demonstrate the feasibility of performing the clinically useful classification from patients of multiple pneumonia types.