Search

Management of data is an increasingly challenging task for enterprises because of the increasingly diverse and complex requirements that come to bear from various fronts. One issue is the protection of the privacy of the information held, while ensuring compliance with various rules and regulations relating to data management.One issue is the protection of the privacy of the information held, while ensuring compliance with various rules and regulations relating to data management. Privacy may be defined as the claim of individuals, groups and institutions to determine for themselves when, how and to what extent information about them is communicated to others. In its broadest scope this is a problem that has challenged civil rights advocates the world over, at different times. [ABSTRACT FROM AUTHOR]

Faced with growing knowledge management needs, enterprises are increasingly realizing the importance of seamlessly integrating critical business information distributed across both structured and unstructured data sources. Academicians have focused on this problem but there still remain a lot of obstacles for its widespread use in practice. One of the key problems is the absence of schema in unstructured text. In this paper we present a new paradigm for integrating information which overcomes this problem – that of Context Oriented Information Integration. The goal is to integrate unstructured data with the structured data present in the enterprise and use the extracted information to generate actionable insights for the enterprise. We present two techniques which enable context oriented information integration and show how they can be used for solving real world problems. [ABSTRACT FROM AUTHOR]

Recently, frequent itemsets mining over data streams attracted much attention. However, mining closed itemsets from data stream has not been well addressed. The main difficulty lies in its high complexity of maintenance aroused by the $\underline{exact}$ model definition of closed itemsets and the dynamic changing of data streams. In data stream scenario, it is sufficient to mining only approximated frequent closed itemsets instead of in full precision. Such a compact but close-enough frequent itemset is called a relaxed frequent closed itemsets. In this paper, we first introduce the concept of $\mathcal{RC}$ (Relaxed frequent Closed Itemsets), which is the generalized form of approximation. We also propose a novel mechanism CLAIM, which stands for CLosed Approximated Itemset Mining, to support efficiently mining of $\mathcal{RC}$. The CLAIM adopts bipartite graph model to store frequent closed itemsets, use Bloom filter based hash function to speed up the update of drifted itemsets, and build a compact HR-tree structure to efficiently maintain the $\mathcal{RC}$s and support mining process. An experimental study is conducted, and the results demonstrate the effectiveness and efficiency of our approach at handling frequent closed itemsets mining for data stream. [ABSTRACT FROM AUTHOR]

Query-adaptive XML indexing has been proposed and shown to be an efficient way to accelerate XML query processing, because it dynamically adapts to the workload. However, existing adaptive index lack of support for branching queries, and also with low efficiency for query processing and adaptation operations. In this paper, we propose a new Adaptive index for Branching queries, which is named as AB-Index. It is designed to support XML path queries with branching predicates. Efficient index construction, query processing, and index adaptation algorithms are proposed for AB-Index. In the experiments, the proposed index is demonstrated to outperform the state-of-the-art approach in the area of adaptive index in terms of query and adaptation efficiencies. [ABSTRACT FROM AUTHOR]

In this paper, we address the problems of adaptive schema mappings between different peers in peer-to-peer networks and searching for interesting data residing at different peers based on such mappings. We begin by classifying the shared schema of each peer into a taxonomy of relation categories and attribute categories. We then propose our adaptive schema mapping by selectively probing the shared schema with query probes, which are generated by the classification rules. To improve the accuracy of schema mapping, we introduce the notion of confusion matrix and prior-knowledge. Finally, we present the query reformulation strategy for retrieving and integrating data from all relevant peers. We have implemented our proposed schema mapping and query processing methods in real settings with real datasets. The experimental results show that our method can be adopted effectively in practice. [ABSTRACT FROM AUTHOR]

RSS (Really Simple Syndication) is a format used for the publication and syndication of web content. While several frameworks, techniques and algorithms have been proposed and studied for the processing of complex queries on data streams, current RSS reader and aggregator software and services do not propose advanced query facilities. We designed and implemented a prototype RSS aggregator service, called Danaïdes, for the processing of complex queries on continuously updated RSS feeds and of progressively producing results. We demonstrate the prototype and its several user-interfaces with a geographical application using geoRSS feeds. This work is a practical application of our research on progressive query processing algorithms for data streams. [ABSTRACT FROM AUTHOR]

This demonstration presents OntoDB, a prototype that allows to store explicitly in the database not only the data, but also the conceptual model defining the structure of data and the domain ontology representing the meaning of data. The demonstration illustrates three main functionalities of OntoDB: (1) a storage of a domain ontology and database content in the same repository, (2) the possibility of querying databases at ontology level, and (3) an automatic integration of heterogeneous data sources referencing/extending the same domain ontology. [ABSTRACT FROM AUTHOR]

The evolution from relational DBMS to data integration system brings new challenges to the design and implementation of query execution engine that must be extended to support queries over multiple distributed, heterogeneous, and autonomous data sources. In this paper, we introduce our work on extending PostgreSQL to support distributed query processing. Although PostgreSQL has no built-in distributed query processor, its function mechanism provides possibilities for us to integrate data of various data sources and execute distributed queries. We point out several limitations in PostgreSQL's query engine and present corresponding query execution techniques to improve performance of distributed query processing. Our experimental results show that the techniques can significantly reduce response times when running a workload consisting of TPC-H queries. [ABSTRACT FROM AUTHOR]

TinTO is an experimental system aiming at demonstrating the usefulness and feasibility of applying conventional SQL queries for analyzing a wide spectrum of data streams. As application area we have chosen the analysis of streams of stock market data, mainly because this kind of application exhibits sufficiently many of those characteristics for which relational query technology can be reasonably considered in a stream context. TinTO is a technical investor tool for computing so-called technical indicators, numerical values calculated from a certain kind of stock market data, characterizing the development of stock prices over a given time period. In contrast to other approaches, TinTO computes indicator values directly over the database by means of SQL queries/views. [ABSTRACT FROM AUTHOR]

This paper focuses on the issue of geographical data's copyrights protection. A Geo-WDBMS has been built by embedding the watermarking functions into the inner code of the open source DBMS PostgreSQL. And its core watermarking mechanism is to insert and detect mark bits in the coordinates of the vertices in geographical objects using the methods of classifying and twice majority-voting. Further more, error correcting mechanism is used to enhance the resilience of the system and blind watermark is realized. Experiments on watermarking digital maps showed that the marked maps are inconspicuous and robust to various attacks. [ABSTRACT FROM AUTHOR]

Recently, several approaches and systems were proposed to store in the same database data and the ontologies describing their meanings. We call these databases, ontology-based databases (OBDBs). Ontology-based data denotes those data that represent ontology individuals (i.e., instance of ontology classes). To speed up query execution on the top of these OBDBs, efficient representations of ontology-based data become a new challenge. Two main representation schemes have been proposed for ontology-based data: vertical and binary representations with a variant called hybrid. In these schemes, each instance is split into a number of tuples. In this paper, we propose a new representation of ontology-based data, called table per class. It consists in associating a table to each ontology class, where all property values of a class instance are represented in a same row. Columns of this table represent those properties of the ontology class that are associated with a value for at least one instance of this class. We present the architecture of our ontology-based databases and a comparison of the effectiveness of our representation scheme with the existing ones used in Semantic Web applications. Our benchmark involves three categories of queries: (1) targeted class queries, where users know the classes they are querying, (2) no targeted class queries, where users do not know the class(es) they are querying, and (3) update queries. [ABSTRACT FROM AUTHOR]

Ubiquitous environment that is omnipresent is existent everywhere or seems to be always present. Such an environment is a next generation paradigm in which many invisible computers are integrated into background of our lives. However, it requires more secure technologies to protect privacy because user may access information without time and space restriction. In this paper, we propose the mechanism that a user is able to automatically access to resource by means of context aware on ubiquitous computing environments. For this purpose, we exploit Role-Based Access Control (RBAC) and Mandatory Access Control (MAC) policies, and defines extended context rules. We also provide an advanced security authorization mechanism and show how to securely preserve properties despite of dynamic change of access control privilege. [ABSTRACT FROM AUTHOR]

The study of integrity constraints has been identified as one of the major challenges in XML database research. The main difficulty is finding a balance between the expressiveness and the existence of automated reasoning tools. We investigate a previous proposal for functional dependencies in XML (XFDs) that is based on homomorphisms between data trees and schema trees. We demonstrate that reasoning about our XFDs is well-founded. We provide a finite axiomatisation and show that their implication is equivalent to the logical implication of propositional Horn clauses and thus decidable in time linear in the size of the constraints. Hence, our XFDs do not only capture valuable semantic information but also permit efficient automated reasoning support. [ABSTRACT FROM AUTHOR]

In Service Oriented Architecture (SOA), the need for inter-service compatibility analysis has gone beyond what existing service composition/ verification approaches can handle. Given two services whose interface invocation constraints are described by Business Process Execution Language for Web Services (BPEL4WS, or BPEL), we analyze their compatibility and adopt mediation as a light weight approach, to make partial compatible services work together more adaptively, without changing their internal logic. We transform BPEL into service workflow net which is a kind of colored Petri net. Based on this formalism we first analyze the compatibility of two services, and then devise an approach to check whether there exists any message mediation so that their mediation-aided composition will not violate the constraints imposed by either side. Later the method for mediation generation is also introduced. Our approach is validated through a real life case and further research directions are pointed out. [ABSTRACT FROM AUTHOR]

This paper addresses the issue of expert finding in a social network. The task of expert finding, as one of the most important research issues in social networks, is aimed at identifying persons with relevant expertise or experience for a given topic. In this paper, we propose a propagation-based approach that takes into consideration of both person local information and network information (e.g. relationships between persons). Experimental results show that our approach can outperform the baseline approach. [ABSTRACT FROM AUTHOR]

We introduce the versioned relational data model, which allows a user to apply conditional schema changes to a populated database without breaking applications compiled against an existing schema, and without loss of existing data. Our model is based on keeping a history of conditional schema changes, and converting tuples on demand to fit the correct schema in any schema version. We provide a concrete definition of schema versioning: The ability to specify an operator on any schema version, such that the tuples in the result are unaffected by schema versions created after the specified schema version. Finally, we show that our model supports schema versioning. [ABSTRACT FROM AUTHOR]

In the past decade, the Internet has rapidly become a most prevalent platform for information sharing and data communications all over the world. This trend has been enhanced by the wide proliferation of home-used digital photos and videos. Although the Web has seen numerous applications for photo storage, sharing, and searching, few image retrieval systems provide satisfactory search service. The so-called semantic gap and curse of dimensionality are two major barriers that existing technologies cannot adequately address. In this paper, we propose a framework to achieve effective and efficient content-base retrieval on images by addressing the above two issues. That is, we need to narrow the semantic gap while mitigating the curse of dimensionality. [ABSTRACT FROM AUTHOR]

In this article, we introduce a continuous data stream reduction method using wavelets summarization. Especially we consider storing a plenty of past data stream into stable storage (flash memory or micro HDD) rather than keeping only recent streaming data allowable in memory, because data stream mining and tracking of past data stream are often required. In the general method using wavelets, a specific amount of streaming data from a sensor is periodically compressed into fixed size and the fixed amount of compressed data (selected wavelet coefficients) is stored into stable storage. However, our method flexibly adjusts the number of selected wavelet coefficients for each local time section. Experimental results with some real world data show that our flexible approach has lower estimation error than the general fixed approach. [ABSTRACT FROM AUTHOR]

Similarity retrieval have been widely used in many practical search applications. A similarity query model can be viewed as a logical combination of a set of similarity predicates. A user can initialize a query model, but model parameters or the model itself may be inadequately specified. As a result, a retrieval system cannot guarantee that it has presented all the relevant tuples to the user. In this paper, we propose a framework that integrates the similarity retrieval and skyline exploration. Using the relevance feedback as a way to constrain the search space, our framework can intelligently explore only a necessary portion of data that contains all the relevant tuples. Our framework is also flexible enough to incorporate model refinement techniques to retrieving relevant results as early as possible. [ABSTRACT FROM AUTHOR]

Query processing in data grids is a complex issue due to the heterogeneous, unpredictable and volatile behavior of grid resources. Considering the existence of multiple partial replicas for each relation and the volatile nature of grid environment, this paper investigates the issues and proposes an adaptive, load-balanced join processing approach. Analytical and experimental results show the effectiveness of our approach. [ABSTRACT FROM AUTHOR]

In this paper, we address a data-level data integration problem where the compared data is semantically equivalent but the data native representation of the values of given attributes is different. We assume that a semantic relationship between potentially used terms is established by a human expert prior to the designed computations, and is represented in an auxiliary table built and maintained for each attribute. We introduce a notion of semantically equal join (SEJ), which is the join operation based on pre-defined semantic relationship. Our goal is to propose a solution for SEJ that can be supported by a standard SQL. [ABSTRACT FROM AUTHOR]

Much research has been done on Keyword Search Over Relational Databases(KSORD) in recent years, and several prototypes have been developed. However, the performance of KSORD systems still is a key issue. In this paper, we propose a combined approach QuickCN for efficient KSORD. Firstly, schema graph is employed to generate Candidate Networks(CNs). Then, data graph is exploited to quickly execute CNs instead of submitting them to RDBMS. In this way, QuickCN performs more efficiently than schema-graph-based KSORD systems and consumes less memory than that by data-graph-based KSORD systems. Our experiments show that QuickCN is efficient and effective. [ABSTRACT FROM AUTHOR]

The previous spatial clustering methods calculate the distance value between two spatial objects using the Euclidean distance function, which cannot reflect the grid path, and their computational complexity is high in the presence of obstacles. Therefore, in this paper, we propose a novel spatial clustering algorithm called DBSCAN-MDO. It reflects the grid path in the real world using the Manhattan distance function and reduces the number of obstacles to be considered by grouping obstacles in accordance with MBR of each cluster and filtering obstacles that do not affect the similarity between spatial objects. [ABSTRACT FROM AUTHOR]

We study a novel problem: tracing data transformations. That is, for a particular target data type, e.g., obtained from the output schema, we trace over the transformation specifications and extract from them the fragments that are exactly used to compute instance data of the type. Our work provides a piecemeal fashion to understand a transformation semantic, and hence would be useful for users to test, debug, and refine the transformation specifications. [ABSTRACT FROM AUTHOR]

General k-anonymity models cannot fully prevent individual re-identification on released microdata since intruders can capture various prior information to recognized individual identities with enough precision, ultimately, to precisely associate these identities with some sensitive attribute values. We propose the privacy inference logic to specify the k-anonymity method and emphasize the nature of privacy inference attacks on k-anonymized microdata in a more rigorous way (than the previous work, including ℓ-diversity), which can be used as a guideline and a predictor for efficient privacy disclosure control during k-anonymization process on a pre-published microdata set. Based on this theory, we uncover and define several main inference attacks classes in aspect of the various "knowledge" used by intruders. In experiments, we successfully implement the probabilistic inference risks evaluation as factors considered in an anonymization cost metric as a more effective anti-inference k-anonymity solution. [ABSTRACT FROM AUTHOR]

Sequential pattern mining is very important because it is the basis of many applications. Although there has been a great deal of effort on sequential pattern mining in recent years, its performance is still far from satisfactory because of two main challenges: large search spaces and the ineffectiveness in handling dense datasets. To offer a solution to the above challenges, we have proposed a series of novel algorithms, called the LAst Position INduction (LAPIN) sequential pattern mining, which is based on the simple idea that the last position of an item, α, is the key to judging whether or not a frequent k-length sequential pattern can be extended to be a frequent (k+1)-length pattern by appending the item α to it. LAPIN can largely reduce the search space during the mining process, and is very effective in mining dense datasets. Our performance study demonstrates that LAPIN outperforms PrefixSpan [4] by up to an order of magnitude on long pattern dense datasets. [ABSTRACT FROM AUTHOR]

As XML information proliferates on the Web, searching XML information via a search engine is crucial to the experience of both casual and experienced Web users. The returned XML fragments in the list is not directly usable, if not confusing, to the users, since in most cases the XML fragments extracted from a large XML repository are incomplete, scattered and redundant. Thus, it is necessary to re-iterate the searching process based on user preferences in order to obtain more complete, detailed and usable results. In this paper, we propose a unifying framework which takes searching, merging and user preferences into account. We view search queries and fragment labeling as an input in an on-going searching process, in which the relevant XML fragments are merged into a concise form and returned to the user a ranked result list. [ABSTRACT FROM AUTHOR]

Nowadays P2P information systems can be considered as large scale databases where all peers can store and query data in the network. Keywords and structure indexes must be maintained. However, indexing XML documents with massive set of words brings out a major problem: The number of entries to be shipped in the network is huge. We define Distributed Bloom Filter, a data structure derived from Bloom Filters, a probabilistic data structure to test whether an element is member of a set, to summarize peer XML content and structure. Our strength is to split the traditional Bloom Filter into several segments. We rely on a DHT network to distribute these segments in a P2P network. Our measurements show that our indexing method is scalable for a large number of words, and outperforms similar methods. [ABSTRACT FROM AUTHOR]

Tree Pattern Queries [7,6] are now well admitted for modeling parts of XML Queries. Actual works only focus on a small subpart of XQuery specifications and are not well adapted for evaluation in a distributed heterogeneous environment. In this paper, we propose the TGV (Tree Graph View) model for XQuery processing. The TGV model extends the Tree Pattern representation in order to make it intuitive, has support for full untyped-XQuery queries, and for optimization and evaluation. Several types of Tree Pattern are manipulated to handle all XQuery requirements. Links between Tree Patterns are called hyperlinks in order to apply transformations on results. The TGV has been implemented in a mediator system called XLive. [ABSTRACT FROM AUTHOR]

In this paper we identify the opportunities for the semantic transformation of XPath queries using the structural and explicit semantics defined in an XML schema. Our classification of transformation is the semantic path expression where a path can be semantically contracted, expanded or complemented. Among several applications of such transformations, an obvious one is the semantic optimization of XPath queries. The transformation is likely to result in an improved response time for a given system. We empirically evaluate the gain or loss of performance of the identified transformations with two representative off-the-shelf XML data management systems and XPath query processors. [ABSTRACT FROM AUTHOR]

OWL is a recommended language for publishing and sharing ontologies on the Semantic Web. To manage the ontologies, several OWL data management systems have been proposed. However, the existing systems have limitations of the scalability and the reasoning. In this paper, we propose an OWL data management system, ONTOMS, which stores OWL data into class based relations, performs complete inverseOf, symmetric, and transitive reasoning for instances, and efficiently evaluates OWL-QL queries against ontologies in a relational database. [ABSTRACT FROM AUTHOR]

Networks of thousands of sensors present a feasible and economic solution to some of our most challenging problems, such as real-time traffic modeling, military sensing and tracking. Many research projects have been conducted by different organizations regarding wireless sensor networks; however, few of them discuss how to estimate missing sensor data. In this research we present a novel data estimation technique based on association rules derived from closed frequent itemsets generated by sensors. Experimental results compared with the existing techniques using real-life sensor data show that closed itemset mining effectively imputes missing values as well as achieves time and space efficiency. [ABSTRACT FROM AUTHOR]

In this paper, we focus on the processing of stream data whose characteristics vary unpredictably by over time. Particularly, we suggest a method which generates an efficient operator execution order called WT-Heuristic. Our method changes the execution order with respect to the change of data characteristics with minimum overheads. [ABSTRACT FROM AUTHOR]

Radio frequency identification (RFID) technology is believed to be the next revolutionary step in supply-chain management. Complex process simplification using RFID technology can offer particularly important benefits to many enterprises. To derive real benefit from RFID, the application must rapidly implement functions to process the large quantity of event data generated by RFID operations. For this reason, developers are forced to implement systems to derive meaningful high-level events from simple RFID events. Although applications could directly consume and act on RFID event, extracting the business rules from the business logic leads to better decoupling of the system, which consequentially, increases maintainability. In this paper, we describe an RFID business aware framework for extending RFID events using business rules, and then processing these to show complex events. [ABSTRACT FROM AUTHOR]

Similarity search over stream time series has a wide spectrum of applications. Most previous work in static time-series databases and stream time series aim at retrieving the exact answer to a similarity search. However, little work considers the approximate similarity search in stream time series. In this paper, we propose a weighted locality-sensitive hashing (WLSH) technique, which is adaptive to characteristics of stream data, to answer approximate similarity search over stream time series. Due to the unique requirement of stream processing, we present an efficient method to update hash functions adaptive to stream data and maintain hash files incrementally at a low cost. Extensive experiments demonstrate the effectiveness of WLSH, as well as the efficiency of approximate similarity search via hashing on stream time series. [ABSTRACT FROM AUTHOR]

To support secure exchanging and sharing of XML data over the Internet, a myriad of XML access control mechanisms have been proposed. In the setting of node-level fine-grained access control, query evaluation is a process of locating XML nodes that (1) satisfy query constraints, and (2) do not violate security policies. In this regard, we propose and empirically validate a suite of XML indices for multi-level XML security model. [ABSTRACT FROM AUTHOR]

Third generation (3G) mobile networks rely on distributed architectures where Operation and Maintenance Centers handle a large amount of information about network behavior. Such data can be processed to extract higher-level knowledge, useful for network management and optimization. In this paper we apply reduction techniques, such as Principal Component Analysis, to identify orthogonal subspaces representing the more interesting data contributing to overall variance and to split them up in "normal" and "anomalous" subspaces. Patterns within anomalous subspaces allow for early detection of network anomalies, improving mobile networks management and reducing the risk of malfunctioning. [ABSTRACT FROM AUTHOR]

Recently a lot of work on integrating the search interfaces of multiple Web databases of the same domain into an integrated interface has been reported. Such integrated interfaces enable users to search multiple Web databases using one query. However, there are two potential problems when using these integrated interfaces in practice. First, if the number of domains is large, it may be difficult for users to find the correct domain. Second, the integrated interfaces can become too complicated for ordinary users to use. In this paper, we propose a system called EasyQuerier to tackle these problems. EasyQuerier allows the users to submit keyword-based queries to access the Web databases by first mapping a keyword-based user query to a suitable domain and then translating the user query to a well-formatted query on the integrated interface of the found domain. Our experiments show that both our domain mapping and query translation techniques work very well. [ABSTRACT FROM AUTHOR]

The study of the Web graph not only yields valuable insight into Web algorithms for crawling, searching and community discovery, and the sociological phenomena that characterize its evolution, but also helps us understand the evolution process of the Web. In this paper, we report the experiments on properties of the Korea Web graph with over 116 million pages and 2.7 billion links. This paper presents the power law distributions from the Korea Web and then compares them with other web graphs. Our analysis reveals that the Korea Web graph has different properties in comparison with the other graphs in terms of the structure of the Web. [ABSTRACT FROM AUTHOR]

Existing approaches to ontology retrieval solely base their search mechanisms on keyword matching while taxonomic structure is solely used for ranking purpose. Users, therefore, are not equipped with expressive means to structurally and semantically describe their ontology needs. To tackle this problem, this paper develops a framework for Semantic Query based Ontology Retrieval, namely SQORE. It enables precise formulation of a semantic query in order to best capture a user' s ontology requirements, which include not only the desired class and property names, but also their relations and restrictions. SQORE employs XML Declarative Description (XDD) theory as its theoretical foundation for modeling ontology databases and evaluating semantic queries. Moreover, similarity score is formally defined as a key metric for ranking the resulting ontologies based on their conceptual closeness to the given query. [ABSTRACT FROM AUTHOR]

In this paper, we propose a new algorithm for information bottleneck method in multi-view setting where instances have multiple independent representations. By introducing the two important conditions, conditional independence and compatibility, into the information bottleneck clustering, the compatible constraint maximizing the agreement between clustering hypotheses on different views is imposed on the individual views to cluster instances. Our algorithm is developed by the compatible constraint. Experiments on three real-world datasets indicate that our algorithm considering the relationship among multiple views can provide solution with improved quality in multi-view setting. [ABSTRACT FROM AUTHOR]