/ Subhash Bhalla / Associate Professor
/ Nadia Berthouze / Assistant Professor
Many types of new data types are commonly used within the web data resources. This year the laboratory organized a workshop on "Databases in Networked Information System (DNIS 2000)" with international participation. Many aspects of web related research activity were discussed at the workshop. In addition to complexity in features such as multiple attributed data, many new types of raw data are emerging that need to be captured by DBMSs for information extraction. Many research efforts are being made to make suitable Object-Relational architectures for spatial databases.
Most of the advances in techniques concentrate on capturing more meaning within data. A number of researchers are actively developing improved data management strategies using XML (extensive markup language) technology. This provides a challenging area for study. The domains of activity include :
Information Processing Systems of future will be a combination of integrated components. There will be components for intelligent problem solving, or decision making, components for specialized data processing and components for shared information management. The applications will utilize a shared base of information. Some examples are - business automation, industrial automation, computer-aided design and manufacture, and cartography.
Traditional approaches to transaction management introduce elements of unpredictable delays during transaction processing. Thus, making these not suitable for adoption in new application environments. The techniques for time-critical transactions are applicable to Mobile databases and multimedia databases.
The goal of this research is to study the feasibility to endow a computational system with the ability to process information. As an experimental platform the laboratory is building a web retrieval system that enables users to retrieve multimedia information carrying a given subjective (Kansei) content. Such system will prove useful in various areas of information technology such as online shopping, advertising industry, entertainment, design, wearable computers. The modeling issues are reformulated from the complementary points of view:
Refereed Journal Papers
Concurrency control based on distributed locking, introduces delays due to local processing and message communication. With improvements in processor speeds, there is a simultaneous increase in contrast, between these delays. In this study, a new approach has been proposed to reduce delays by enhancing local processing, and reducing the inter-site communication delays. The proposal presents an efficient synchronization algorithm for exchange of precedence in case of a conflict. The technique is efficient in global time-stamp management, and incurs a minimal overall synchronization overhead.
Databases support transaction update activities on continuou s and on-going bases. Reading an entire database is needed for many applications. Existing techniques introduce blocking. These restrict the transaction update activity. We propose an algorithm that reads current database entities without interference with update activity. The algorithm is simple to implement as compared with previous proposals. It assigns a color to each entity read by the global-read. Normal transactions commit by declaring a color for the committed updates. Color markings are used to update and for generation of a consistent copy of the entire database, in an off-line manner.
In a real-time database system, an application supports a mix of transactions. These include the real-time transactions that require completion by a given deadline. Time-critical requirements also exist in many distributed multi-media system applications. Existing concurrency control procedures introduce excessive delays due to non-availability of data resources. In this study, we ignore the delays incurred by ordinary transactions, in order to achieve a non-interference mode of execution (near parallel) for the time-critical transactions. For this purpose, a data allocation model has been studied. It is a stochastic process model based on the use of two-phase locking.
This is a position paper that identifies the features requir ed for systems to be capable of capturing affective information. Those features are: embodiment (experiencing physical reality), dynamics (mapping experience and emotional state with its label) and adaptive interaction (conveying emotive response, responding to a recognize emotional state). The paper extensively defines affect and emotion in the view of affective computing. The discussion is supported by two complementary systems developed by the authors. A first system recognizes the emotional state of the user by interpreting his/her visual expression and subsequently develops semantic definitions of emotion concepts through user feedback. The second system co-evolves with the user a language for communicating over subjective impression. This co-evolution relies on externalization process triggered by the system and aimed to lead the user to explicit his/her internal model of his/her own subjectivity.
Refereed Proceeding Papers
To recover from media failures, a database is 'restored' from an earlier backup copy. A recovery log of transactions is used to roll forward from the backup version to the desired time (the current time). High availability requires - backup activity to be fast, and on-line with on-going update activity. It necessitates, obtaining a consistent copy of an entire database. Such concurrent generation of a database copy, interferes with system activity. It introduces blocking and delays for many update transactions. We study the performance of revised algorithms, to highlight the level of concurrent activity permitted by these algorithms, in parallel.
With the growing importance of information technology in today's industry, new types of applications are appearing that require a broader understanding of information, including subjective parameters, for applications in design industry, advertisement, entertainment or internet-shopping. We propose here our framework for endowing a software agent with the capability to personalize itself to its user through interaction and perform tasks that involve subjective parameters. We describe K-Dime, a software prototype that can retrieve material from the Web on the basis of both objective and subjective features of the content. K-Dime allows users to create their own Kansei User Model. With its ability to bootstrap a new user model from the model of a user with similar profile, it significantly reduces the workload generally associated with an online learning phase. The user can also be involved in the creation of his/her kansei model through interactive session with the software agent. Continuous adaptation driven by specific patterns of interaction with the user enables K-Dime to cope with the intrinsic variability of subjective impressions. A working prototype has been implemented and evaluation results are discussed.
In this paper, we address the issue of reducing the complexity and uncertainty of the mapping between patterns in images and impression (kansei) words, i.e. two expression modalities for a same visual impression. The approach we propose relies on the active role to be played by the user in the modeling process. His/her role is to externalize his/her mental model of the mapping. A conceptual space is proposed as a formal space where user and system can derive formal hypotheses from the externalization process of the user. The conceptual space is based on a set of predefined categories defined on low-level characteristics of images and a set of evolving categories (impression words being modeled). The dimensions in the low-level category are geometrically and spatially specified, the higher level categories are built as a combination of the lower one. Three main mechanisms allow the user to construct his/her mental process and are identified as: inconsistency detection, similarity measurement and word specialization. In order to support the user in this difficult activity, cognitive maps are used to visually and objectively represent the paths in the process. We believe that the approach proposed here can be extended to other modalities (music, body language, etc.) and whenever subjective imopressions should be conveyed or understood.
Data management systems for large applications seek infra structural support from database systems and distributed computing systems. These systems improve with the changing requirements of applications. New research efforts are being made to support Electronic Commerce (EC) applications, and voluminous data handling applications. At the same time, support techniques are being studied in data mining and knowledge discovery, object-relational database management systems, and mobile computing systems. The DNIS 2000 research activity has been organized to to emphasize the research activity on support systems.