In a Big Data era, a lot of open data set is published and shared with the public. That creates new services and business. However, the publication may cause a leakage problem of private information. In general, de-identification techniques are applied to the data before publication. The problem, however, has not been solved completely. Personal data can be obtained from the several sources such as Internet service and social media. In this situation, a de-identified open data may be simply joined with the leaked external data and it may result in a re-identification issue. We propose a new analytic model to measure the personal information leakage risk in the open data before publishing. The proposed model formulates the entropy-based re-identification risk to measure the privacy leakage risk. We also try to find the data utility measure by using the entropy while preserving the privacy. Based on both the risk and the utility measure, we propose the guideline for data open to the public. We show the guideline including the risk and utility measurement can be applicable with the empirical experiments.

The volume of XML data is tremendous in many areas, but especially in data logging and scientific areas. XML data in the areas are accumulated over time as new data are continuously collected. It is a challenge to process massive XML data with multiple twig pattern queries given by multiple users in a timely manner. We showcase HadoopXML, a system that simultaneously processes many twig pattern queries for a massive volume of XML data with Hadoop. Specifically, HadoopXML provides an efficient way to process a single large XML file in parallel. It processes multiple twig pattern queries simultaneously with a shared input scan. Users do not need to iterate M/R jobs for each query. HadoopXML also reduces many I/Os by enabling twig pattern queries to share their path solutions each other. Moreover, HadoopXML provides a sophisticated runtime load balancing scheme for fairly assigning multiple twig pattern joins across nodes. With synthetic and real world XML dataset, we demonstrate how efficiently HadoopXML processes many twig pattern queries in a shared and balanced way.

There are multiple data sources storing avian influenza virus information in Korea. Research on AI requires scientists to collect, integrate, share and analyze AI virus information from the data sources. Since the sources are heterogeneous, autonomously created and maintained by different institutions, it is very difficult for AI researchers to search their data sets from the sources. In this paper, we propose a mediator-based data integration system, which integrates such multiple heterogeneous AI virus information sources. Mediation approach to integrating the sources requires scientists to define the semantic mapping among them. Most AI researchers do not have enough knowledge on the sources to define the mapping. To alleviate their efforts, AVIS provides a visual administration tool to browse schema of data sources and easily define the semantic mapping for data integration. It also provides Web-based graphical interface to easily access and share data from multiple heterogeneous AI virus information sources.

Conventional search engines generally cannot find information from the Deep Web because they use hyper link-based crawling techniques to visit Web pages. Recently, lots of research efforts are being tried to crawl the Deep Web. One of the obstacles for crawling the Deep Web is the requirement of huge computing resources, but most of search engine companies hardly meet the needs. We, therefore, propose the design of the Grid-based middleware, OGSA-DWC for crawling the Deep Web. With our middleware, developers will easily implement a Grid-based Deep Web crawling system although they do not have much knowledge about how to use idle and distributed computing resources.

This paper presents a system architecture and information retrieval strategy for multimedia content which exploits descriptive metadata as well as domain ontology. We propose a query processing model including a semantic ranking scheme which can retrieve multimedia objects semantically relevant to the user query and provide users with a search result categorized by concepts and ordered by their semantic relevance to the query. By experiments on a prototype system using a repository of broadcast TV programs, we show effectiveness of the ontology-based multimedia search approach proposed in the paper.

This article proposes a novel video abstraction framework for online review services of story-oriented videos such as dramas. Among the many genres of TV programs, a drama is one of the most popularly watched on the Web. The abstracts generated by the proposed framework not only give a summary of a video but also effectively help viewers understand the overall story. In addition, our method is duration-flexible. We get clues about human understanding of a story from scenario writing rules and editorial techniques that are popularly used in the process of video production to explicitly express a narrative, and propose a new video abstraction model, called a Narrative Abstraction Model . The model effectively captures the narrative structure embedded in a story-oriented video and articulates the progress of the story in a weighted directed graph, called a Narrative Structure Graph (NSG) . The model provides a basis for a flexible framework for abstract generation using the NSG as the intermediary representation of a video. Different abstracts can be appropriately generated based upon different user requirements. To show the effectiveness of the proposed model and method, we developed a video abstraction system realizing the framework, and successfully applied it to large volumes of TV dramas. The evaluation results show that the proposed framework is a feasible solution for online review services.

TV program review services, especially drama review services, are one of the most popular video on demand services on the Web. In this paper, we propose a novel video abstraction model for a review service of story-oriented video such as dramas. In a drama review service, viewers want to understand the story in a short time and service providers want to provide video abstracts at minimum cost. The proposed model enables the automatic creation of a video abstract that still allows viewers to understand the overall story of the source video. Also, the model has a flexible structure so that the duration of an abstract can be adjusted depending on the requirements given by viewers. We get clues for human understanding of a story from scenario writing rules and editorial techniques which are popularly used in the process of video producing. We have implemented the proposed model and successfully applied it to several TV dramas.

This paper proposes a cache management scheme for continuous partial match queries in mobile computing systems. Conventional cache management methods for mobile clients are record ID-based ones. However, since the partial match query is a content-based retrieval, the conventional record ID-based approach cannot properly manage the cache consistency. We show the pr edicate- based approach is an effective cache management in mobile environments.

OLAP queries involve a lot of aggregations on a large amount of data in data warehouses. To process expensive OLAP queries efficiently, we propose a new method for rewriting a given OLAP query using the various kinds of materialized aggregate views which already exist in data warehouses. We first define the normal forms of OLAP queries and materialized views based on the lattice of dimension hierarchies and the semantic information in data warehouses. Conditions for the usability of a materialized view in rewriting a given query are specified by relationships between the components of their normal forms. We present a rewriting algorithm for OLAP queries that effectively utilizes existing materialized views. The proposed algorithm can make use of materialized views having different selection granularities, selection regions and aggregation granularities together, to generate an efficient rewritten query.

As the B-tree index could be a performance bottleneck in database systems, concurrent B-tree algorithms have been intensively studied to improve concurrency of B-tree accesses in the literature. We propose a new concurrent B/sup +/-tree algorithm for high concurrency and an efficient tree restructuring method. Because the proposed method of tree restructuring always preserves a semantic consistency of the B/sup +/-tree, a key searcher need not require any lock for a range search as well as a single-key search. To maintain correctly the link fields constructed at the leaf level, we make each leaf node contain two key-range indicators, and develop a tree restructuring method using these key-range indicators.

Data replication is often considered in distributed database systems to enhance availability and performance. The benefit of data replication, however can only be realized at the cost of maintaining the consistency of data. In particular, network partition failures make it more difficult to achieve high data availability while ensuring strong correctness criteria such as l-copy serializability. We propose a replica control method to improve the availability of data in the presence of network partition failures. Our method extends the traditional primary copy method by using the relaxed correctness criterion called insular consistency for large-scale distributed systems, where partition failures frequently occur. We focus on increasing the availability of data for read-only transactions. We introduce a version vector as a tool for guaranteeing insular consistency and present a mechanism that allows read-only transactions to be executed at any partition as long as the insular consistency is satisfied. An asynchronous update propagation mechanism is also employed to improve the performance of update operations. We also show that the proposed method is correct and give some performance considerations.

This paper presents a simple and effective method to reduce the size of log data for recovery in main memory databases. Fuzzy checkpointing is known to be very efficient in main memory databases due to asynchronous backup activities. By this feature, most recovery works in the past have used only physical logging schemes. Since the size of physical log records is quite large, physical logging schemes cause the recovery time to be much longer than that using logical logging schemes. We propose a hybrid logging method that can accommodate logical logging under fuzzy checkpointing. This method significantly decreases the size of log data, and hence speeds up the recovery time. We also propose a reapplying rule in segmented MMDB, which reduces the number of log records for recovery. We evaluate the performance of the proposed methods through analytic analyses. The results show that we can reduce the size of log data to more than half, compared with those that use only physical logging.

With rapid increase of information requirements from various application areas, there has been much research on dynamic information storage structures that effectively support insertions, deletions and updates. We evaluate the performance of the existing dynamic signature file methods such as the S-tree, Quid Filter and HS file, and provide guidelines for the most effective usage to a given operational environment. We derive analytic performance evaluation models of the storage structures based on retrieval time, storage overhead and insertion time. We also perform extensive experiments with various data distributions such as uniform, normal and exponential distributions.

The retrieval models based on the extended boolean retrieval framework, e.g., the fuzzy set model and the extended boolean model have been proposed in the past to provide the conventional boolean retrieval system with the document ranking facility. However, due to undesirable properties of evaluation formulas for the AND and OR operations, the former generates incorrect ranked output in certain cases and the latter suffers from the complexity of computation. There have been a variety of fuzzy operators to replace the evaluation formulas. In this paper we first investigate the behavioral aspects of the fuzzy operators and address important issues to affect retrieval effectiveness. We then define an operator class called positively compensatory operators giving high retrieval effectiveness, and present a pair of positively compensatory operators providing high retrieval efficiency as well as high retrieval effectiveness. All the claims are justified through experiments.

The authors address an enforcement technique of integrity constraints against transaction updates in a relational database system. Transition axioms have been used effectively in checking integrity constraints for single update statements. The authors extend the idea of transition axioms to a transaction which is a sequence of read and update statements. Integrity constraints in this scheme are simplified without database access before the actual operations are performed avoiding the need to undo an illegal transaction. Transaction partitioning, which can reduce the overhead of checking integrity constraints significantly, is proposed. Partitioning a transaction becomes crucial when a transaction is associated with multiple integrity constraints.>

An active database system reacts to a set of external events such as a timer interrupt or access to a particular object in the database. A trigger is a general mechanism for active data management, both in the context of a centralized system or a distributed system (including that of autonomous and cooperating agents). It consists of three parts; event specification, integrity constraint specification, and action specification. The event specification in a trigger is a set of events which will cause the condition in the constraint specification to be checked. If the condition is true, the actions in the action specification will be initiated. In this paper, we develop a framework for supporting triggers in object-oriented database systems. The framework consists of a categorization for each of the three components of a trigger. The framework is first cast in the context of the relational model of data, and is then extended to account for object-oriented concepts that constitute an object-oriented data model, including nested objects, methods, and inheritance hierarchy.

A management discipline of software metrics facilitates their consistent measurement and usage during the life cycle of software systems, by maintaining knowledge and data related to their evaluation tasks and application domain knowledge. To investigate environment of the management discipline, we represent software engineering environments by an activity flow model and a micro cycle model. A unified management view of software metrics is established by their collectivity and comprehensiveness in two-dimensional flows of the activity flow model. Then, we propose the Software Quality Manager (SQM) as a realization of the management discipline of software metrics. The proposed software quality manager makes use of two knowledge bases: (1) knowledge about application domain in the form of reusable software components and (2) knowledge about software engineering. The three-stage architecture and preparation process of the proposed system are also presented. Finally, we propose how to support the measurement and analysis task for database management systems within the context of the software quality manager.

In order to improve the two-level signature file method designed by Sacks-Davis et al. [20], we propose new multikey access methods based on term discrimination and signature clustering. By term discrimination, we create separate, efficient access methods for the terms frequently used in user queries. We in addition cluster similar signatures by means of these terms so that we may achieve good performance on retrieval. Meanwhile we provide the space-time analysis of the proposed methods and compare them with the two-level signature file method. We show that the proposed methods achieve 15-30% savings in retrieval time and require 3-9 % more storage overhead.