Search

This paper presents a novel application of a semantic annotation system, named Cerno, to analyze research publications in electronic format. Specifically, we address the problem of providing automatic support for authors who need to deal with large volumes of research documents. To this end, we have developed Biblio, a user-friendly tool based on Cerno. The tool directs the user's attention to the most important elements of the papers and provides assistance by generating automatically a list of references and an annotated bibliography given a collection of published research articles. The tool performance has been evaluated on a set of papers and preliminary evaluation results are promising. The backend of Biblio uses a standard relational database to store the results. [ABSTRACT FROM AUTHOR]

Some of the most interesting developments within computer system design in recent years have emerged from an exploration of the ways everyday objects and artefacts can be augmented with computational resources. Often under the rubric of "ubiquitous computing", research programmes in Europe, North America and Japan have directed substantial funding towards these initiatives, and leading industrial and academic research laboratories have developed a diverse range of ubiquitous computing "solutions". These developments mark an important shift in system design, a shift that is having a corresponding impact on social science research. Surprisingly though, given the growing commitment to the ubiquitous and the tangible, there is a mundane, even humble artefact that pervades our ordinary lives that has received less attention than one might imagine. This artefact is paper. [ABSTRACT FROM AUTHOR]

The aim of this paper is to explore new metaphors for interaction design on tabletop system. Tabletop systems are shared horizontal surface for co-located collaboration, which leads to original problems when designing interactions. We propose two metaphors based on the paper: the peeling metaphor, and the slot metaphor, and then suggest a way of using them to design new interactions for solving some of the problems of tabletop systems: documents organization, documents transmission and documents duplication. [ABSTRACT FROM AUTHOR]

Many design activities depend on communicative activities around collaboratively produced prototypes. A common communication practice in producing text documents is to add annotation in the form of comments. Previous research indicates that electronic paper-prototyping can be used to rapidly create simple prototypes of interactive systems, such as websites. Little is known, however, about how to provide and maintain variety of communication channels around such electronic paper-prototypes to enable end-users and other stakeholders to contribute to design dialogues. This paper presents Gabbeh, an electronic paper-prototyping tool, and reports on an evaluation using the tool in a simulated design exercise. [ABSTRACT FROM AUTHOR]

Researchers from the same lab often spend a considerable amount of time searching for published articles relevant to their current project. Despite having similar interests, they conduct independent, time consuming searches. While they may share the results afterwards, they are unable to leverage previous search results during the search process. We propose a research paper recommender system that avoids such time consuming searches by augmenting existing search engines with recommendations based on previous searches performed by others in the lab. Most existing recommender systems were developed for commercial domains with millions of users. The research paper domain has relatively few users compared to the large number of online research papers. The two major challenges with this type of data are the large number of dimensions and the sparseness of the data. The novel contribution of the paper is a scalable subspace clustering algorithm (SCuBASCuBA: Subspace Clustering Based Analysis.) that tackles these problems. Both synthetic and benchmark datasets are used to evaluate the clustering algorithm and to demonstrate that it performs better than the traditional collaborative filtering approaches when recommending research papers. [ABSTRACT FROM AUTHOR]

Multiple business factors have emerged to accelerate the necessity of event-driven functionality and make it part of the main-stream computing, instead of a niche technology. Consequently, there is now focus on using high-level software constructs to build these applications. This paper presents a vision for such high-level features and architecture. This paper explains why "event-driven applications" becomes an emerging area, explains the basic terminology of EDA, explains the relationship to business rules, and sets some directions for the future of this discipline. [ABSTRACT FROM AUTHOR]

Office workers tend to produce paper piles of documents to read or to process sometime later. The information contained in these piles is often lost if it is not transferred to electronic format and connected to knowledge structures. Information that is not part of the knowledge worker's electronic information space is frequently overlooked because it is not proactively provided during actual processes or tasks he is involved in. This paper presents a novel prototype for an intelligent office appliance, which results from an integration of three state-of-the-art office applications/appliances: a workflow system, a document classification system, and a multi-functional peripheral. The resulting system allows for leveraging an office worker's papers to her personal knowledge space in order to realize a pro-active and context-sensitive information support within knowledge-intensive tasks and processes. [ABSTRACT FROM AUTHOR]

To facilitate the transition from paper-based patient record (PPR) to computer-based patient record (CPR), engineers should try to improve the usability of CPR system. From the point of view of a physician, we reviewed PPRs written by 8 Japanese physicians. We revealed the characteristics of PPR to find out about the barrier to transition from PPR to CPR. Our findings may be helpful to the engineers who are aiming to develop a CPR system. [ABSTRACT FROM AUTHOR]

In this paper we describe the design of a hybrid search that combines simple metadata search with a traditional keyword search over unstructured context data. This paradigm provides the inquirer additional options to narrow the search with some semantic aspects through the XML metadata query. We demonstrate a paper search for a case study of the hybrid search, and describe a data integration hub to integrate those data dispersed on the Net. [ABSTRACT FROM AUTHOR]

Global paper industry needs systems that can be used in all locations. International user studies can be helpful when integrating systems. Not only due to the lack of common language but also the differences in culture and the usage of systems, information from different countries needs to be collected and analyzed so that the integration development is not be biased and unilateral. During the study some food for thought was gathered on what to consider when planning an international user study. [ABSTRACT FROM AUTHOR]

With the increased use of Internet and information technology all over the world, there is an increased amount of criminal activities that involve computing and digital data. These digital crimes (e-crimes) impose new challenges on prevention, detection, investigation, and prosecution of the corresponding offences. Computer forensics (also known as cyberforensics) is an emerging research area that applies computer investigation and analysis techniques to help detection of these crimes and gathering of digital evidence suitable for presentation in courts. This new area combines the knowledge of information technology, forensics science, and law and gives rise to a number of interesting and challenging problems related to computer security and cryptography that are yet to be solved. In this paper, we present and discuss some of these problems together with two successful cases of computer forensics technology developed in Hong Kong that enable the law enforcement departments to detect and investigate digital crimes more efficiently and effectively. We believe that computer forensics research is an important area in applying security and computer knowledge to build a better society. [ABSTRACT FROM AUTHOR]

Privacy-preserving classification mining is one of the fast-growing sub-areas of data mining. How to perturb original data and then build a decision tree based on perturbed data is the key research challenge. By applying transition probability matrix this paper proposes a novel privacy-preserving classification mining algorithm which suits all data types, arbitrary probability distribution of original data, and perturbing all attributes (including label attribute). Experimental results demonstrate that decision tree built using this algorithm on perturbed data has comparable classifying accuracy to decision tree built using un-privacy-preserving algorithm on original data. [ABSTRACT FROM AUTHOR]

Stochastic Opponent Modeling Agents (SOMA) have been proposed as a paradigm for reasoning about cultural groups, terror groups, and other socioeconomic- political-military organizations worldwide. In this paper, we describe a case study that shows how SOMA was used to model the behavior of the terrorist organization, Hezbollah. Our team, consisting of a mix of computer scientists, policy experts, and political scientists, were able to understand new facts about Hezbollah of which even seasoned Hezbollah experts may not have been aware. This paper briefly overviews SOMA rules, explains how more than 14,000 SOMA rules for Hezbollah were automatically derived, and then describes a few key findings about Hezbollah, enabled by this framework. [ABSTRACT FROM AUTHOR]

Stochastic Opponent Modeling Agents (SOMA) have been proposed as a paradigm for reasoning about cultural groups, terror groups, and other socioeconomic-political-military organizations worldwide. In this paper, we describe the SOMA Terror Organization Portal (STOP). STOP provides a single point of contact through which analysts may access data about terror groups world wide. In order to analyze this data, SOMA provides three major components: the SOMA Extraction Engine (SEE), the SOMA Adversarial Forecast Engine (SAFE), and the SOMA Analyst NEtwork (SANE) that allows analysts to find other analysts doing similar work, share findings with them, and let consensus findings emerge. This paper describes the STOP framework. [ABSTRACT FROM AUTHOR]

Copyright of Signal Processing for Image Enhancement & Multimedia Processing is the property of Springer eBooks and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Information retrieval techniques have been used for a long time to identify links between textual items for the automatic construction of hypertexts and electronic books where sought information can be accessed by browsing. While research work in this area has been steadily decreasing in recent years, some of the techniques developed in that context are proving very valuable in a number of new application areas. In this paper we present an approach to automatic linking of textual items that is used to prioritise criminal suspects in a police investigation. A free-text description of an unsolved crime is compared to previous offence descriptions where the offender is known. By linking the descriptions, inferences about likely suspects can be made. Language Modeling is adapted to produce a Bayesian model which assigns a probability to each suspect. An empirical study showed that the linking of free text descriptions of burglaries enables prioritisation of offenders. The model presented in this paper could be easily extended to take account of additional crime and suspect linking data, such as geographical location of crimes or suspect social networks. This would enable large networks of investigative information automatically constructed from police archives to be browsed. [ABSTRACT FROM AUTHOR]

This paper discusses the recognition of textual entailment in a text-hypothesis pair by applying a wide variety of lexical measures. We consider that the entailment phenomenon can be tackled from three general levels: lexical, syntactic and semantic. The main goals of this research are to deal with this phenomenon from a lexical point of view, and achieve high results considering only such kind of knowledge. To accomplish this, the information provided by the lexical measures is used as a set of features for a Support Vector Machine which will decide if the entailment relation is produced. A study of the most relevant features and a comparison with the best state-of-the-art textual entailment systems is exposed throughout the paper. Finally, the system has been evaluated using the Second PASCAL Recognising Textual Entailment Challenge data and evaluation methodology, obtaining an accuracy rate of 61.88%. [ABSTRACT FROM AUTHOR]

This paper presents a computational framework for information extraction and aggregation which aims to integrate and organize the data/information resources that spread throughout the Internet in the manner that makes them useful for tracking events such as natural disaster, and disease dispersion. We introduce a simple statistical information extraction technique for summarizing the document into a predefined structure. We apply the topic maps approach as a semantic layer in aggregating and organizing the extracted information for smart access. In addition, this paper also carries out a case study on disease dispersion domain using the proposed framework. [ABSTRACT FROM AUTHOR]

Self-organizing map can be an effective tool for the textual data classification. In this paper, we represent the methodology of an integration of the information system modeling and the development of the information system natural language interface. The main idea of the paper is to build the set of self-organising maps from information system documentation and then reuse it in human-machine communication as a semantic parsing component. The IBM's Information Framework (IFW) Financial Services Data Model has been used in an experiment where we tested how appropriate is presented methodology and what is classification accuracy of the received self-organizing maps. We compare classification accuracy with the IBM's WebSphere Voice Server NLU solution and demonstrate that self-organising maps can be a competitive components in the information systems natural language interfaces. [ABSTRACT FROM AUTHOR]

Repositories of knowledge about the real world and how it functions are needed to advance research in intelligent, knowledge-intensive systems. The repositories are intended to serve as surrogates for the meaning and context of terms and concepts. These are being developed at two levels: 1) individual domain ontologies that capture concepts about a particular application domain, and 2) upper level ontologies that contain massive amounts of knowledge about the real world and are domain independent. This paper analyzes ResearchCyc, which is a version of the most extensive base of common sense knowledge, the upper level ontology, Cyc. It does so to summarize the current state of the art in upper level ontology development in order to suggest areas for future research. The paper also describes various problems encountered in applying ResearchCyc to web query processing. [ABSTRACT FROM AUTHOR]

This paper describes a car traffic control simulation realised in a decentralised way by message propagations: congested nodes (roads intersections) send speed-up or slow-down messages to neighbouring nodes. Different types of journeys have been modelled: regular car journeys, accidents and emergency cars journeys. These journeys have different lengths and speeds, and affect the system differently. Optimal values of parameters, used during the simulations for controlling the cars, have been determined through the use of a genetic algorithm (GA). This paper reports as well a preliminary experiment on different simulations realised with parameters values derived from the GA. [ABSTRACT FROM AUTHOR]

Large organizations are increasingly confronted with the problem of capturing, processing, and archiving large amounts of data. For several reasons, the problem is especially cumbersome in the case where data is stored on paper. First, the weight, volume, and relative fragility of paper incur problems in handling and require specific, labor-intensive processes to be applied. Second, for automatic processing, the information contained on the pages must be digitized, performing Optical Character Recognition (OCR). This leads to a certain number of errors in the data retrieved from paper. Third, the identities of individual documents become blurred. In a stack of paper, the boundaries between documents are lost, or at least obscured to a large degree.1 [ABSTRACT FROM AUTHOR]

This paper describes a theoretical framework for inducing knowledge from incomplete data sets. The general framework can be used with any formalism based on a lattice structure. It is illustrated within two formalisms: the attribute-value formalism and Sowa's conceptual graphs. The induction engine is based on a non-supervised algorithm called default clustering which uses the concept of stereotype and the new notion of default subsumption, inspired by the default logic theory. A validation using artificial data sets and an application concerning the extraction of stereotypes from newspaper articles are given at the end of the paper. [ABSTRACT FROM AUTHOR]

It is widely recognized that ontologies can be used to support the semantic integration and interoperability of heterogeneous information systems. Resource Event Agent (REA) is a well-known business ontology that was proposed for ontology-driven enterprise system development. However, the current specification is neither sufficiently explicit nor formal, and thus difficult to operationalize for use in ontology-driven business information systems. In this paper REA is redesigned and formalized following a methodology based on the reengineering extension of the METHONTOLOGY framework for ontology development. The redesign is focused on developing a UML representation of REA that improves upon existing representations and that can easily be transformed into a formal representation. The formal representation of REA is developed in OWL. The paper discusses the choices made in redesigning REA and in transforming REA's UML representation into a OWL representation. [ABSTRACT FROM AUTHOR]

It is recognized that conceptual models expressed in ORM are more suitable for analysis stage and the relational database design, and because of natural verbalization are better tolerated by stakeholders, whereas UML models are more often used in the design of object oriented systems. If the system requires both of these characteristics in particular relational database and object oriented implementation of business logic, the problem of transformation from ORM to UML arises. This paper propose the approach to transform between two well-known modelling techniques: ORM models are transformed into UML models constrained by Object Constraint Language (OCL). The approach precisely describes properties of the transformation. This opens the approach for seamless refining of resulted models using UML tools and transformation to executable code. The transformation of ORM to UML/OCL transformation is validated for correctness by means of a widely used UML tools. Paper illustrates the proposed approach by a number of representative examples. [ABSTRACT FROM AUTHOR]

This paper presents nexus, an event extraction system, developed at the Joint Research Center of the European Commission utilized for populating violent incident knowledge bases. It automatically extracts security-related facts from on-line news articles. In particular, the paper focuses on a novel bootstrapping algorithm for weakly supervised acquisition of extraction patterns from clustered news, cluster-level information fusion and pattern specification language. Finally, a preliminary evaluation of nexus on real-world data is given which revealed acceptable precision and a strong application potential. [ABSTRACT FROM AUTHOR]

There have been many studies on the design of P2P systems for effective keyword search. This paper proposes and tests technique to reduce network traffic due to many inverted lists forwarded in carrying out query containing multi-keywords in DHT base structured p2p network. Many of inverted lists, forwarded inter-medium, are discarded regardless of search result. This paper proposes Distance and Smart-bloom filter to diminish those unrelated inverted lists. Distance can correctly distinguish document not containing a certain keyword. Smart-bloom filter with low false positive rate can sort out document with a high chance of including certain keyword among inverted lists selected by distance. Ultimately large amount of unrelated inverted lists can be diminished. The performance of Distance and Smart-bloom filter was tested through simulation and the traffic was decreased by 67%. [ABSTRACT FROM AUTHOR]

The paper proposes the framework for investigation of prospective adaptive and cooperative defense mechanisms against the Internet attacks. The approach suggested is based on the multi-agent modeling and simulation. According to the approach the defense and attack systems are represented as interacting teams of intelligent agents that act under some adaptation criterion. They adjust their configuration and behavior in compliance with the network conditions and attack (defense) severity. The paper represents the architecture and software implementation of simulation environment that combines discrete-event simulation, multi-agent approach and packet-level simulation of various Internet protocols. The environment allows to simulate complex attack and defense scenarios. The paper describes the experiments aimed on the investigation of adaptive "Distributed Denial of Service" attacks and defense mechanisms. [ABSTRACT FROM AUTHOR]

Distributed decision making and learning procedures operate with distributed data sources. Intelligent access to distributed data sources is one of the important requirements to any distributed learning and decision making systems. In many such applications, input data arrive from spatially distributed sensors structured in sensor networks. An increasing interest to what is called intelligent sensors for which agent-based technology seems to be rather attractive implementation paradigm forms a marked trend in the sensor network area. Indeed, the latter provides a natural mapping from the set of nodes of intelligent sensor network to the set of collaborating agents and, at the same time, enriches the network nodes by the methods of collaboration developed under multi-agent approach and provides the network nodes with the capability to communicate in terms of high level language. The paper proposes a typical (reusable) architecture of the intelligent sensor agents as well as typical protocols supporting interaction of sensors with other software components of agent-based network intended for distributed learning and classification. The main solutions proposed in the paper are demonstrated via prototyping of two case studies. [ABSTRACT FROM AUTHOR]

As modern applications tend to stretch between large, ever-growing datasets and increasing demand for meaningful content at the user end, more elaborate and sophisticated knowledge extraction technologies are needed. Towards this direction, the inherently contradicting technologies of deductive software agents and inductive data mining have been integrated, in order to address knowledge intensive problems. However, there exists no generalized evaluation methodology for assessing the efficiency of such applications. On the one hand, existing data mining evaluation methods focus only on algorithmic precision, ignoring overall system performance issues. On the other hand, existing systems evaluation techniques are insufficient, as the emergent intelligent behavior of agents introduce unpredictable factors of performance. In this paper, we present a generalized methodology for performance evaluation of intelligent agents that employ knowledge models produced through data mining. The proposed methodology consists of concise steps for selecting appropriate metrics, defining measurement methodologies and aggregating the measured performance indicators into thorough system characterizations. The paper concludes with a demonstration of the proposed methodology to a real world application, in the Supply Chain Management domain. [ABSTRACT FROM AUTHOR]

Peer-to-Peer (P2P) networks are gaining increasing popularity in many distributed applications such as file-sharing, network storage, web caching, sear- ching and indexing of relevant documents and P2P network-threat analysis. Many of these applications require scalable analysis of data over a P2P network. This paper starts by offering a brief overview of distributed data mining applications and algorithms for P2P environments. Next it discusses some of the privacy concerns with P2P data mining and points out the problems of existing privacy-preserving multi-party data mining techniques. It further points out that most of the nice assumptions of these existing privacy preserving techniques fall apart in real-life applications of privacy-preserving distributed data mining (PPDM). The paper offers a more realistic formulation of the PPDM problem as a multi-party game and points out some recent results. [ABSTRACT FROM AUTHOR]

This paper proposes an interconnection approach, which is based on extended many-valued context and extended formal descriptions. An extended many-valued context ∏=(G,M,Q,W,I) consists of sets G,M,Q and W and a quaternary-relation I $\subseteq$G×M×Q×W. An extendedformal description is regarded as a mapping from the set of attributes to the power set of the values, assigning to each attribute the set of allowed values under some conditions. The extended formal descriptions are naturally ordered by preciseness, and then a concept lattice is obtained according to the theory of FCA. This concept lattice is the well structure interconnection among concepts. The paper also proposed some important propositions, which are used to decide whether two concepts have semantic interconnections. In the end, the paper describes an interconnection algorithm with the time complexity O(n2). [ABSTRACT FROM AUTHOR]

An advanced fuzzy C-mean(FCM) algorithm for the efficient regional clustering of multi-nodes interconnected systems is presented in this paper. Owing to physical characteristics of the interconnected systems, nodes or points in the interconnected systems have their own information indicating the network-related characteristics of the system. However, classification for the whole system into distinct several subsystems based on a similarity measure is typically needed for the efficient operation of the whole system. In this paper, therefore, a new regional clustering algorithm for interconnected systems based on the modified FCM is proposed. Moreover, the regional information on the system are taken into account in order to properly address the geometric mis-clustering problem such as grouping geometrically distant nodes with similar measures into a common cluster. We have presented that the proposed algorithm has produced proper classification for the interconnected system and the results are demonstrated in the example of IEEE 39-bus interconnected electricity system. [ABSTRACT FROM AUTHOR]

Although Data Field Clustering method has a lot of advantages, clustering result depends severely on affection factor that is selected in Data Field function. The purpose of the paper is to find an optimum affection factor that may not only reflect nature characteristic of clustering data sample, but also reduce influence caused by sample deviation to minimum. In this paper, an affection interval concept is defined at first. Then an optimum objective function for reducing influence of sample deviation is constructed and an approximate solution is given of optimum affection factor. In the end, a standard data set offered in the MATLAB is used to test the availability of the optimum affection factor, the result is satisfactory. [ABSTRACT FROM AUTHOR]

L2 and L1 constrained regression methods, such as ridge regression and Lasso, have been generally known for their fitting ability. Recently, L0-constrained classifications have been used for feature selection and classifier construction. This paper proposes an L0 constrained regression method, which aims to minimize both the epsilon-insensitive fitting errors and L0 constraints on regression coefficients. Our L0-constrained regression can be efficiently approximated by successive linearization algorithm, and shows the favorable properties of selecting a compact set of fitting coefficients and tolerating small fitting errors. To make our L0 constrained regression generally applicable, the extension to nonlinear regression is also addressed in this paper. [ABSTRACT FROM AUTHOR]

This paper presents a new clustering technique which is extended from the technique of clustering based on frequent-itemsets. Clustering based on frequent-itemsets has been used only in the domain of text documents and it does not consider frequency levels, which are the different levels of frequency of items in a data set. Our approach considers frequency levels together with frequent-itemsets. This new technique was applied in the domain of bio-informatics, specifically to obtain clusters of genes of zebrafish (Danio rerio) based on Expressed Sequence Tags (EST) that make up the genes. Since a particular EST is typically associated with only one gene, ESTs were first classified in to a set of classes based on their features. Then these EST classes were used in clustering genes. Further, an attempt was made to verify the quality of the clusters using gene ontology data. This paper presents the results of this application of clustering based on frequent-itemsets and frequency levels and discusses other domains in which it has potential uses. [ABSTRACT FROM AUTHOR]

Principal curves were proposed as the nonlinear generalization of PCA. However, for the tasks of feature extraction for signal representation at which PCA is adept, existing definitions of principal curves have some weakness in their theoretical bases thus fail to get reasonable results in many situations. In this paper, a new definition of principal curves - Principal Curve with Feature Continuity (PCFC) is proposed. PCFC focuses on both reconstruction error minimization and feature continuity. It builds a continuous mapping from samples to the extracted features so the features preserve the inner structures of the sample set, which benefits the researchers to learn the properties of the sample set. The existence and the differential properties of PCFC are studied and the results are presented in this paper. [ABSTRACT FROM AUTHOR]

The availability of large volumes of text documents has created the potential of a vast amount of valuable information buried in those texts. This in turn has created the need for automated methods of discovering relevant information without having to read it all. This paper focuses on detecting links between two concepts across text documents. We interpret such a query as finding the most meaningful evidence trail across documents that connect these two concepts. In this paper we propose to use link-analysis techniques over the extracted features provided by Information Extraction Engine for finding new knowledge. We compare two approaches to perform this task. One is the concept-profile approach based on traditional bag-of-words model, and the other is the graph-based approach which combines text mining, graph mining and link analysis techniques. Counterterrorism corpus is used to evaluate the performance of each model and demonstrates that the graph-based approach is preferable for finding focused information. For greater coverage of information we should use the concept-profile based approach. [ABSTRACT FROM AUTHOR]

Use of relevance feedback (RF) in the feature vector model has been one of the most popular approaches for fine tuning query for content-based image retrieval (CBIR) systems. This paper proposes a framework that extends the RF approach to capture the inter-query relationship between current and previous queries. By using the feature vector model, this approach avoids the need of "memorizing" actual retrieval relationship between the actual image indexes and the previous queries. This implies that the approach is more suitable for image database application where images are frequently added or removed. This paper has extended the authors' previous work [1] by applying a semantic structure to connect the previous queries both visually and semantically. In addition, active learning strategy has been used in this paper to explore images that may be semantically similar while visually different. [ABSTRACT FROM AUTHOR]

This paper is concerned with the problem of mining commodity information from threaded Chinese customer reviews. Chinese online commodity forums, which are developing rapidly, provide a good environment for customers to share reviews. However, due to noises and navigational limitations, it is hard to have a clear view of a commodity from thousands of related reviews. Further more, due to different characters between Chinese and English, Researching approaches may vary a lot. This paper aims to automatically mine out key information from commodity reviews. An effective algorithm, i.e. Chinese Commodity Review Miner (CCRM) is proposed. The algorithm can be divided into two parts. First, we propose an efficient rule based algorithm for commodity feature extraction as well as a probabilistic model for feature ranking. Second, we propose a top-to-down algorithm to reorganize the extracted features into hierarchical structure. A prototype system based on CCRM is also implemented. Using CCRM, users can easily acquire the outline of a commodity, and navigate freely in it. [ABSTRACT FROM AUTHOR]

This paper proposes an approach to detect duplicates among relational data. Traditional methods for record linkage or duplicate detection work on a set of records which have no explicit relations with each other. These records can be formatted into a single database table for processing. However, there are situations that records from different sources can not be flattened into one table and records within one source have certain (semantic) relations between them. The duplicate detection issue of these relational data records/instances can be dealt with by formatting them into several tables and applying traditional methods to each table. However, as the relations among the original data records are ignored, this approach generates poor or inconsistent results. This paper analyzes the characteristics of relational data and proposes a particular clustering approach to perform duplicate detection. This approach incorporates constraint rules derived from the characteristics of relational data and therefore yields better and more consistent results, which are revealed by our experiments. [ABSTRACT FROM AUTHOR]