Next: Information Systems Laboratory Up: Department of Computer Previous: Performance Evaluation Laboratory

Database Systems Laboratory


/ Subhash Bhalla / Associate Professor

Existing research activity in the area of database systems aims to capture a wide variety of applications into the fold of DBMS (Database Management System) technology. This is due to improved understanding of information extracted from databases by the usage of advanced DBMS techniques. The DBMSs currently use visual representations within new interfaces to manage large volumes of data, and also to manage data that may have complex features associated with its structure. In addition to complexity in features, such as multiple attributed data, wave-forms and multi-dimensional orientation, many new types of raw data are emerging that need to be captured by DBMSs for the purpose of information extraction. Among these, the image data and video image data are two important forms of data and many research efforts are being made to make suitable DBMS architectures for image and multimedia databases.

In order to provide flexibility in terms of data representation, the databases need the support of powerful data-modeling techniques. Most of the advances in techniques concentrate on capturing more meaning within data that is made resident on a computer system. Within this framework, a number of researchers are actively developing improved data management strategies.

On the implementation side, existing database systems consist of many layers of software within a DBMS ( Database Management System). These layers are mostly independent of each other. The layers are prepared to meet a specific function, such as parallel accesses or transaction commit. These impose many delays and processing constraints. In addition to this fact, at many times the software architecture is an old design, because software systems are slower to develop as compared to hardware design changes. Therefore, the DBMS architectures are based on inefficient design techniques, in many cases. Due to these two factors, many design bottlenecks exist that lead to lower level of performance of database systems, especially when dealing with many new database applications, concerning with industrial applications.

A new set of performance factors are important in new time-critical applications such as in Mobile databases for electronic commerce, and Multimedia applications using Real-time databases, in process and manufacturing industries. Thus, our aim is to reduce the worst time estimates and increase dependability of computation within a specified unit of time. Many research efforts are being made to develop capabilities within the next generation software systems that need to support time-critical transactions. This provides a challenging area for study in many domains of activity. These include :

  1. Next generation Operating Systems for distributed and multi-media applications;
  2. Current mission critical applications in Process Control, such as in Electrical Power Grid Control, and in manufacturing industries;
  3. Mobile computing and databases for electronic commerce;
  4. Large Network Management Database Management Systems, for Giga-byte networking projects; and
  5. Multi-media Synchronization applications

A high level of reliability of operations is also neccesary within these systems to ensure quick response in case of failure. Thus, the activity of performance evaluation of transaction processing systems is an important area that provides improved understanding of systems. This facilitates improved integration and refinement in technique. The Database Systems Laboratory has research activity concentrating on data modeling as well as transaction processing activity. A brief outline of the ongoing projects is given below.


Refereed Proceeding Papers

  1. S. Bhalla, H. Abramson, K. Christianson, J.M. Goodwin, J.R.Goodwin, J. Sarraille, On-line Japanese Character and Word Recognition Using Databases of Identifiable Features of Characters and Compounds. 5th International Conference on Japanese Information in Science, Technology, and Commerce, p.121-126, US Department of commerce, and US library of congress, July 1997.

    The study of Japanese as a second language is made more difficult by the fact that learning to read is a more time consuming task. The difficulty is compounded by the lack of suitable reference materials. Currently available dictionaries have two major drawbacks: these are not comprehensive enough to serve all users. Also, these do not offer easy and rapid methods of kanji lookup. The dictionary using our ( proposed ) method can take advantage of the virtually unlimited space in electronic media to include all the knaji and kanji compounds.

  2. Bhalla, S., H. Abramson, K. Christianson, J.M. Goodwin, J.R.Goodwin, J. Sarraille, L.M. Schmitt, Recognition of Japanese Kanji Characters by Non-Japanese Learner's Through a Support Database System. 2nd International Conference on Cognitive Technology, p.190-199, IEEE Computer Society, August 1997.

    Our paper presents a qualitative analysis of differences in the way that kanji are perceived by different users of Japanese language. Most non-native users of Japanese dictionaries find it difficult to use these dictionaries, because the data are organized in the conventional way, that is similar to dictionaries for native Japanese users. These present compilations are based on the assumption that each user has sufficient knowledge of the language. However, the non-native users, in particular at the beginner's level, have a number of difficulties concerning the `kanji', and words.

  3. L.M. Schmitt, J. Herder, S. Bhalla, Information Retrieval and Database Architecture for Conventional Japanese Character Dictionaries. 2nd International Conference on Cognitive Technology, p.200-217, IEEE Computer Society, August 1997.

    The method of abstract modeling and usage of the model is described in this paper. This leads to detection and development of many cognitive tools. This shown that as a part of the process, a conventional Japanese - English dictionary can be converted as a more convenient tool as a multidimensional database for easier reference. The methods used in the design of the multidimensional database include the use of UNIX software tools, and classical database methods.

  4. Shreenivas, M.V., and Bhalla, S., Independent Global Snapshots in Large Distributed Systems. 4th International Conference on High Performance Computing, p.462-467, ACM SigArch, IEEE Computer Society, December 1997.

    Distributed systems depend on consistent global snapshots for process recovery and garbage collection activity. We provide exact conditions for an arbitrary checkpoint based on independent dependency tracking within clusters of nodes. The method permits that nodes (within clusters) can independently compute dependency information based on available ( local ) information. The existing models of global snapshot computations provide the necessary and sufficient conditions. But, these require expensive global computations. The nodes can also compute conditions to make a checkpoint, or conditions, such that a collection of checkpoints, can belong to a global snapshot.

Unrefereed Papers

  1. Bhalla, S., Reducing Data Contention for Transaction Processing within a Mobile Database System. Advanced Database Systems for Integration of Media and User Environments, Vol. 9. p.99-103, World Scientific Publishers, February 1998.

Grants

  1. Sabhash Bhalla, Ministry of Education Scientific Research Fund, High Priority Research (A), Computer Science, Database Systems, grant number 09230217.

Others

  1. Ookawara, H., Minimizing Power Consumption in Mobile Computing Systems. The Univ. of Aizu, 1998. Thesis Advisor: Sabhash Bhalla.

  2. Narahashi M., Communication and Consistency in Mobile Data Systems. The Univ. of Aizu, 1998. Thesis Advisor: Sabhash Bhalla.

  3. Hiroyuki Yui, Algorithms for High Performance in Real-time Database Systems. The Univ. of Aizu, 1998. Thesis Advisor: Sabhash Bhalla.

  4. Nomura M., Privacy and Authentication in Mobile ata Access Systems. The Univ. of Aizu, 1998. Thesis Advisor: Sabhash Bhalla.



Next: Information Systems Laboratory Up: Department of Computer Previous: Performance Evaluation Laboratory


www@u-aizu.ac.jp
December 1998