Cheng-Hsien Tang

Actor garbage collection using vertex-preserving actor-to-object graph transformations

Large-scale distributed computing applications require concurrent programming models that support modular and compositional software development The actor model supports the development of independent software components with its asynchronous message-passing communication and state encapsulation properties Automatic actor garbage collection is necessary for high-level actor-oriented programming, but identifying live actors is not as intuitive and easy as identifying live passive objects in a reference graph However, a transformation method can turn an actor reference graph into a passive object reference graph, which enables the use of passive object garbage collection algorithms and simplifies the problem of actor garbage collection In this paper, we formally define potential communication by introducing two binary relations - the may-talk-to and the may-transitively-talk-to relations, which are then used to define the set of live actors We also devise two vertex-preserving transformation methods to transform an actor reference graph into a passive object reference graph We provide correctness proofs for the proposed algorithms The experimental results also show that the proposed algorithms are efficient.

Enterprise Email Classification Based on Social Network Features

With the popularity of multimedia and network technologies, it is now often to attach large size of multimedia dataset to emails. However, delivering large volume of multimedia data over an enterprise email system can easily bring down the quality of overall network service. Moreover, without some sort of restrictions, many enterprises found that the network resource was occupied for personal interests. The business communication over emails thus suffers undesirable delays and cause damages to businesses. The competition to use email service therefore become an issue that many enterprises have to deal with. Obviously, enterprises should manage the email service so that business emails have the priority over personal usages. This management requires an effective methodology to classify enterprise emails into official and private emails, and the development of the method is the goal of this work. To achieve the accuracy of a desired classification methodology, we normally anticipated the developed method to survey as much information as possible. On the other hand, monitoring details of the email contents not only can decrease the performance of the method, but it also may violate the privacy rights that many legal regulation systems now protect. The balance of pursuing accurate classification and protecting privacy rights becomes a challenge for this problem. With the discussed challenges in mind, we develop an email classification method based on social features, rather than surveying the email contents. To the best of our knowledge, this paper is the first study to address the aforementioned problems. We obtain social features from emails to represent the input vector of support vector machine (SVM) classifier. Preliminary results show that our methodology can classify emails with a high accuracy. Compared with the other content-based feature of email, our work shows that exploring social features is a promising direction to solve similar email classification problems.

Social feature-based enterprise email classification without examining email contents

Without imposing restrictions, many enterprises find nonwork-related contents consuming network resources. Business communication over emails thus incurs undesired delays and inflicts damages to businesses, explaining why many enterprises are concerned with the competition to use email services. Obviously, enterprises should prioritize business emails over personal ones in their email service. Therefore, previous works present content-based classification methods to categorize enterprise emails into business or personal correspondence. Accuracy of these methods is largely determined by their ability to survey as much information as possible. However, in addition to decreasing the performance of these methods, monitoring the details of email contents may violate privacy rights that are under legal protection, requiring a careful balance of accurately classifying enterprise emails and protecting privacy rights. The proposed email classification method is thus based on social features rather than a survey of emails contents. Social-based metrics are also designed to characterize emails as social features; the obtained features are treated as an input of machine learning-based classifiers for email classification. Experimental results demonstrate the high accuracy of the proposed method in classifying emails. In contrast with other content-based methods that examine email contents, the emphasis on social features in the proposed method is a promising alternative for solving similar email classification problems.

Service Mining for Composite Service Discovery

Web service technology is being applied to organizing business process in many large-scale enterprises. Discovery of composite service, therefore, has become an active research area. In this paper, we utilize a PLWAP-tree algorithm to analyze the relationship among web services from web service usage log. This method generates time-ordered sets of web services which can be exploited to integrate into a real business process. The empirical result shows the methodology is useful, flexible, and efficient. It is able to integrate web services into a composite service according to the mining result.

Hierarchical role classification based on social behavior analysis

Role identification is a difficult job for many social network applications. One of the difficulties is to maintain and utilize large amount of distinct roles. In this paper, we attempt to adopt fuzzy classification method and construct a hierarchy for role classification. We believe this approach can encourage the utilization of social roles by considering their identifiable features at different levels.

An efficient distributed hierarchical-clustering algorithm for large scale data

The data-classification process can possibly involve a huge amount of data in todays cloud computing environment. It could take a long time for processing, and could consume many resources for computation and storage. This study focuses on the problem of using the traditional hierarchical agglomerative clustering algorithm on a distributed environment since hierarchical agglomerative clustering has high applicability and efficiency. A parallel hierarchical ag-glomerative clustering algorithm is proposed in this study. The proposed algorithm divides the whole computation into several small tasks, distribute the tasks to message-passing processes, and merge the results to form a hierarchical cluster. A threshold is used to reduce the storage requirement during the computation. To evaluate the performance and limitation of our algorithm, this study has conducted several experiments using real astronomical data, the main asteroid belt catalog. The experimental results confirm that the proposed parallel algorithm is efficient.

Sequential pattern discovery for Intrusion Detection System

Intrusion Detection System (IDS) is the key technology to ensure the security of dynamic systems. We employ a sequential pattern mining approach to discover significant system call sequences to prevent malicious attacks. To reduce the computing time of generating meaningful rules, we design a weighted suffix tree structure to detect intrusive events on the fly. The experimental results show our method can substantially enhance the accuracy and efficiency of IDS.

Constructing Storage Capacity Migration Policies for Information Lifecycle Management System

To maintain very large amount of increasing dataset, it is critical to enforce the data migration policies across different levels of storage devices. The challenge is to achieve both low storage cost and response efficiency. In this paper, we will investigate the factors for this problem, then adapt dynamic programming technique, and design appropriate ECA trigger rules. From the conducted experiments, our methodology provides a feasible solution for this important kernel issue of the Information Lifecycle Management (ILM) problem.

Shortest-linkage-based parallel hierarchical clustering on main-belt moving objects of the solar system

Data clustering is an important data preparation process in many scientific analysis researches. In astronomy, although the distributed environments and modern observation techniques enable users to collect and access huge amounts of data, the corresponding clustering process may become very costly. One of the challenges is that the sequential clustering algorithms, that can be applied to cluster hundreds of thousand main-belt asteroids to reason about the origins of the main-belt asteroids, may not be used in the distributed environment directly. Therefore, this study focuses on the problem of parallelizing the traditional hierarchical agglomerative clustering algorithm using shortest-linkage. We propose a new parallel hierarchical agglomerative clustering algorithm based on the master-worker model. The master process divides the whole computation into several small tasks, and distributes the tasks to the worker processes for parallel processing. Then, the master process merges the results from the worker processes to form a hierarchical data structure. The proposed algorithm uses a pruning threshold to reduce the execution time and the storage requirement during the computation. It also supports fast incremental update that merges new data items into a constructed hierarchical tree in seconds, given a tree of about 550,000 data items. To evaluate the performance of our algorithm, this study has conducted several experiments using the MPCORB dataset and a dataset from the DVO database. The results confirm the efficiency of our proposed methodology. Compared with prior similar studies, the proposed algorithm is more flexible and practical in the problem of distributed hierarchical agglomerative clustering.

Efficient astronomical data classification on large-scale distributed systems

Classification of different kinds of space objects plays an important role in many astronomy areas Nowadays the classification process can possibly involve a huge amount of data It could take a long time for processing and demand many resources for computation and storage In addition, it may also take much effort to train a qualified expert who needs to have both the astronomy domain knowledge and the capability to manipulate the data This research intends to provide an efficient, scalable classification system for astronomy research We implement a dynamic classification framework and system using support vector machines (SVMs) The proposed system is based on a large-scale, distributed storage environment, on which scientists can design their analysis processes in a more abstract manner, instead of an awkward and time-consuming approach which searches and collects related subset of data from the huge data set The experimental results confirm that our system is scalable and efficient.