Showing papers by "William W. L. Cheung published in 2005"

Automatic generation of artistic chinese calligraphy

Abstract: Chinese calligraphy is among the finest and most important of all Chinese art forms and an inseparable part of Chinese history. Its delicate aesthetic effects are generally considered to be unique among all calligraphic arts. Its subtle power is integral to traditional Chinese painting. A novel intelligent system uses a constraint-based analogous-reasoning process to automatically generate original Chinese calligraphy that meets visually aesthetic requirements. We propose an intelligent system that can automatically create novel, aesthetically appealing Chinese calligraphy from a few training examples of existing calligraphic styles. To demonstrate the proposed methodology's feasibility, we have implemented a prototype system that automatically generates new Chinese calligraphic art from a small training set.

Learning the kernel matrix for XML document clustering

Abstract: The rapid growth of XML adoption has urged for the need of a proper representation for semi-structured documents, where the document structural information has to be taken into account so as to support more precise document analysis. In this paper, an XML document representation named "structured link vector model" is adopted, with a kernel matrix included for modeling the similarity between XML elements. Our formulation allows individual XML elements to have their own weighted contribution to the overall document similarity while at the same time allows the between-element similarity to be captured. An iterative algorithm is derived to learn the kernel matrix. For performance evaluation, the ACM SIGMOD record dataset as well as the CEDE dataset have been tested. Our proposed method outperforms significantly the traditional vector space model and the edit-distance based methods. In addition, the kernel matrix obtained as a by-product provides knowledge about the conceptual relationship between the XML elements.

On knowledge grid and grid intelligence: a survey

Abstract: The next generation Web Intelligence (WI) aims at enabling users to go beyond the existing online information search and knowledge queries functionalities and to gain, from the Web,1 practical wisdom for problem solving. To support such a Wisdom Web, we envision that a grid-like computing infrastructure with intelligent service agencies is needed, where these agencies can interact, self-organize, learn, and evolve their course of actions, identities, and interrelationships for new knowledge creation, as well as scientific and social evolution. In this paper, we first provide an overview of recent development in WI and Semantic/Knowledge Grid. Then, the fundamental capabilities of the Wisdom Web as well as the conceptual architecture of an intelligent Grid for supporting it are described. Technical challenges for realizing Grid Intelligence are highlighted and the recent advancements in related research areas are reviewed. 1.1. Web Intelligence and Wisdom Web The Web has irrevocably revolutionized the world we live in. This impact is inevitable due to the facts that the Web connectivity rapidly increases and that the online information astronomically explodes. In order not only to live with such a change but also to benefit from the information infrastructure that the Web has empowered, we have witnessed the fast development as well as applications of many Web Intelligence (WI) techniques and technologies (Zhong, Liu, and Yao 2003), which cover: 1. Internet-level communication, infrastructure, and security protocols. The Web is regarded as a computer-networked system. WI techniques for this level include, for instance, Web data-prefetching systems built upon Web-surfing patterns to resolve the issue of Web latency. The intelligence of the Web prefetching comes from adaptive learning based on observations of user-surfing behavior. 2. Interface-level multimedia presentation standards. The Web is regarded as an interface for human‐Internet interaction. WI techniques for this level are used to develop the intelligent Web interfaces in which the capabilities of adaptive cross-language processing, personalized multimedia representation, and multimodel data processing are required. 3. Knowledge-level information processing and management tools. The Web is regarded as a distributed data/knowledge base. We need to develop semantic markup languages to represent the semantic contents of the Web available in machine-understandable formats for agent-based computing, such as searching, aggregation, classification, filtering, managing, mining, and discovery on the Web (Berners-Lee, Hendler, and Lassila 2001). 4. Application-level ubiquitous computing and social intelligence environments. The Web is regarded as a basis for establishing social networks that contain communities for establishing social networks that contain communities of people (or organizations or other social entities) connected by social relationships, such as friendship, coworking, or information exchange with common interests. They are Web-supported social networks or virtual communities. The study of WI concerns the important issues central to social

Integrating Element and Term Semantics for Similarity-Based XML Document Clustering

Abstract: Structured link vector model (SLVM} is a recently proposed document representation that takes into account both structural and semantic information for measuring XML document similarity. Its formulation includes an element similarity matrix for capturing the semantic similarity between XML elements - the structural components of XML documents. In this paper, instead of applying heuristics to define the similarity matrix, we proposed to learn the matrix using pair-wise similar training data in an iterative manner. In addition, we extended SLVM to SLVM-LSI by incorporating term semantics into SL VM using latent semantic indexing, with the element similarity related properties of the original SLVM preserved. For performance evaluation, we applied SLVM-LSI to similarity-based clustering af two XMZ. datasets and the proposed SLVM-LSI was found to significant(y outpeform the conventional vector space model and the edit-distance based methods. The similarity matrix. obtained as a by-product via the learning, can provide higher-level knowledge about the semantic relationship between the XML elements.

Learning global models based on distributed data abstractions

Abstract: Due to the increasing demand of massive and distributed data analysis, achieving highly accurate global data analysis results with local data privacy preserved becomes an increasingly important research issue. In this paper, we propose to adopt a model-based method (Gaussian mixture model) for local data abstraction and aggregate the local model parameters for learning global models. To support global model learning based on solely local GMM parameters instead of virtual data generated from the aggregated local model, a novel EM-like algorithm is derived. Experiments have been performed using synthetic datasets and the proposed method was demonstrated to be able to achieve the global model accuracy comparable to that of using the data regeneration approach at a much lower computational cost.

State of the Art in Simulating Future Changes in Ecosystem Services

Abstract: Coordinating Lead Author: Peter Kareiva Lead Authors: John B. R. Agard, Jacqueline Alder, Elena Bennett, Colin Butler, Steve Carpenter, W. W. L. Cheung, Graeme S. Cumming, Ruth Defries, Bert de Vries, Robert E. Dickinson, Andrew Dobson, Jonathan A. Foley, Jacqueline Geoghegan, Beth Holland, Pavel Kabat, Juan Keymer, Axel Kleidon, David Lodge, Steven M. Manson, Jacquie McGlade, Hal Mooney, Ana M. Parma, Miguel A. Pascual, Henrique M. Pereira, Mark Rosegrant, Claudia Ringler, Osvaldo E. Sala, B. L. Turner II, Detlef van Vuuren, Diana H. Wall, Paul Wilkinson, Volkmar Wolters Review Editors: Robin Reid, Marten Scheffer, Antonio Alonso

Visualizing global manifold based on distributed local data abstractions

Abstract: Mining distributed data for global knowledge is getting more attention recently. The problem is especially challenging when data sharing is prohibited due to local constraints like limited bandwidth and data privacy. In this paper, we investigate how to derive the embedded manifold (as a 2-D map) for a horizontally partitioned data set, where data cannot be shared among the partitions directly. We propose a model-based approach which computes hierarchical local data abstractions, aggregates the abstractions, and finally learns a global generative model - generative topographic mapping (GTM) based on the aggregated data abstraction. We applied the proposed method to two benchmarking data sets and demonstrated that the accuracy of the derived manifold can effectively be controlled by adjusting the data granularity level of the adopted local abstraction.

Grounding collaborative knowledge building in semantics-based critiquing

Abstract: In this paper we investigate the use of Latent Semantic Analysis (LSA), Critiquing Systems, and Knowledge Building to support computer-based teaching of English composition. We have built and tested an English Composition Critiquing System that make use of LSA to analyze student essays and compute feedback by comparing their essays with teacher's model essays. LSA values are input to a critiquing component to provide a user interface for the students. A software agent can also use the critic feedback to coordinate a collaborative knowledge building session with multiple users (students and teachers). Shared feedback provides seed questions that can trigger discussion and extended reflection about the next phase of writing. We present the first version of a prototype we have built, and report the results from an informal experiment. We end the paper by describing our plans for future work.

Dynamic sub-ontology evolution for collaborative problem solving

Abstract: Emerging Web technologies have enabled the use of distributed on-line resources to support on-demand cooperative problem solving. Domain ontology is one of the resources typically required by many existing problemsolving systems. In this paper, we argue that to meet the ondemand requirement, on-line problem-solvers should go beyond the use of static domain ontology and be able to self-evolve and specialize in the knowledge they possess (called sub-ontology). Building on top of the Semantic Web technology, we propose an agent-oriented architecture and a sub-ontology evolution mechanism for dynamic selforganization of domain sub-ontologies to result in a truly intelligent and on-demand problem-solving platform in a distributed environment like the Web/Grid.

Towards end-to-end privacy control in the outsourcing of marketing activities: a web service integration solution

Abstract: With the recent adoption of marketing activities outsourcing, there have been increasing demands and concerns for privacy control. The traditional approach of a bulk transmission of the customers' information to a marketing company cannot meet such demands, especially in the finance and healthcare businesses. Therefore, we propose a layered architecture and a development methodology for end-to-end privacy control over the export of each individual customer's records through a Web services platform, according to the corresponding enterprise's privacy control policies. A Web services system, with up-dated security and privacy facilities, can provide a suitable interoperation platform for required application-to-application interactions over the Internet. We further develop a conceptual model and an interaction protocol to send only the required part of a customer's records at a time. We illustrate our approach for end-to-end privacy control with a tele-marketing case study and show how the software of the outsourced call center can be integrated effectively with the Web services of a bank to protect privacy.

Decomposing large-scale POMDP via belief state analysis

Abstract: Partially observable Markov decision process (POMDP) is commonly used to model a stochastic environment with unobservable states for supporting optimal decision making Computing the optimal policy for a large-scale POMDP is known to be intractable Belief compression, being an approximate solution, has recently been proposed to reduce the dimension of POMDP's belief state space and shown to be effective in improving the problem tractability In this paper, with the conjecture that temporally close belief states could be characterized by a lower intrinsic dimension, we propose a spatio-temporal brief clustering that considers both the belief states' spatial (in the belief space) and temporal similarities, as well as incorporate it into the belief compression algorithm The proposed clustering results in belief state clusters as sub-POMDPs of much lower dimension so as to be distributed to a set of distributed agents for collaborative problem solving The proposed method has been tested using a synthesized navigation problem (Hallway2) and empirically shown to be able to result in policies of superior long-term rewards when compared with those based on solely belief compression Some future research directions for extending this belief state analysis approach are also included

Sub-ontology evolution for service composition with application to distributed e-learning

Abstract: In order to meet the on-demand requirement of service composition at a large scale, one should go beyond the use of static domain ontologies but allow different focused aspects of the ontologies to be distributed as sub-ontologies To demonstrate the feasibility of the sub-ontology idea, we suggest a possible implementation using the semantic Web technology and apply it to distributed e-learning

A critiquing system to support English composition through the use of latent semantic analysis

Abstract: Many students find essay writing stressful because they do not have sufficient ideas to fully cover the topic of the essay. They usually run out of ideas before they can complete their essays. In this paper, we investigate the use of critiquing systems [G. Fischer et al., (1991), J.E. Robbins, (1998)] and latent semantic analysis [T.K. Landauer et al., (1998), D.J. Steinhart, (2001)] to support computer-supported teaching of English composition at the high school level. We start by characterizing essay composition as a design activity and make use of Donald Schon's [D.A. Schon, (1983)] theory of reflection-inaction to identify key elements of an English composition design environment. We make use of latent semantic analysis (LSA) to analyze student essays and compute feedback by comparing their essays with teacher's model essays. We describe the basics of LSA, explain a system architecture that incorporates LSA, present the first version of a prototype we have built, and report the results from an experiment. We end the paper by describing some of our plans for future work.