The Dowdy and Foster paper, "Comparative Models of the File Assignment Problem," COMPUTING SURVEY... more The Dowdy and Foster paper, "Comparative Models of the File Assignment Problem," COMPUTING SURVEYS, 14, 2 (June 1982), 287-313, discussed the optimal distribution of files among storage nodes in a computer system. Kollias and Hatzopoulos write to describe some additional models and studies of the file assignment problem-Ed.
Using parallelism and pipeline for the optimisation of join queries
Lecture Notes in Computer Science, 1992
Abstract: In this study we present a technique for the parallel optimisation of join queries, tha... more Abstract: In this study we present a technique for the parallel optimisation of join queries, that uses the offered coarse-grain parallelism of the underlying architecture in order to reduce the CPU-bound optimisation overhead. The optimisation technique performs an almost ...
A Cost Model for the Estimation Query Execution Time in a Parallel Environment Supporting Pipeline
Computing and Informatics / Computers and Artificial Intelligence, 1996
We propose a model for the estimation of query execution time in an environment supporting bushy ... more We propose a model for the estimation of query execution time in an environment supporting bushy and pipelined parallelism. We consider a parallel architecture of processes having private main memories, accessing a shared secondary storage and communicating to each other via a network. For this environment, we compute the cost of query operators when processed in isolation and when in pipeline mode. WE use those formulae to incrementally compute the cost of a query execution plan from its components. Our cost model can be incorporated to any optimizer for parallel query processing that considers parallel and pipelined execution of the query operators.
Agents for Querying Distributed Statistical Databases Over the Internet
International Journal on Artificial Intelligence Tools, 2002
Distributed database techniques and the Internet provide producers of statistics with a means to ... more Distributed database techniques and the Internet provide producers of statistics with a means to publish their data and metadata widely and make them available to a variety of users. Data matching to a user query and data access as well as data harmonization are some of the problems that should be solved. Intelligence is required in various stages of query answering and data matching. Moreover, the breadth and distributed nature of the Internet urge for a distributed approach. Agents seem to be the means by which both intelligence and distributed processing can be achieved. This paper presents a distributed approach for answering queries on statistical data that exist over the Internet using a multi-agent framework.
Parallel optimization of large join queries with set operators and aggregates in a parallel environment supporting pipeline
IEEE Transactions on Knowledge and Data Engineering, 1996
We propose a parallel optimizer for queries containing a large number of joins, as well as set op... more We propose a parallel optimizer for queries containing a large number of joins, as well as set operators and aggregate functions. The platform of execution is a shared-disk multiprocessor machine supporting bushy parallelism and pipeline. Our model partitions the query into almost independent subtrees that can be optimized simultaneously and applies an enhanced variation of the iterative improvement technique on
The main issues under consideration when facing network and distributed computing stem from the i... more The main issues under consideration when facing network and distributed computing stem from the interconnection of various elements residing in different sites over a local or wide area network (LAN or WAN). Ways of making possible the communication between different processes being executed on different computers have to be developed, a practice that is widely known under the term interprocess communication (IPC). These communication mechanisms are nothing more than abstractions, mainly software, enabling processes to ...
There are two main objectives of the DIAS (European Digital Upper Atmosphere Server) project. Fir... more There are two main objectives of the DIAS (European Digital Upper Atmosphere Server) project. First, it establishes a pan-European repository of raw and derived digital data describing the state of ionospheric part of the upper atmosphere, which is capable of ingesting real-time information and maintaining historical data collections provided by most operating ionospheric stations in Europe. Second, the DIAS system produces and distributes, based on the raw data collection, several operational products required by various user groups for ionospheric nowcasting, prediction and forecasting purposes. The project completed on May 2006 and the DIAS server operates since then continuously. The basic products that are delivered are real-time and historical ionograms from all DIAS-affiliated ionospheric stations, frequency plots and maps of the ionosphere over Europe based on the foF2, M(3000)F2, MUF and electron density parameters, as well as long term prediction and short term forecasting...
Parallel Optimisation of Join Queries Using a Technique of Exhaustive Nature 1
Abstract. In this study we present a technique for the parallel optimisation of join queries that... more Abstract. In this study we present a technique for the parallel optimisation of join queries that uses the offered coarse-grain parallelism of the underlying architecture in order to reduce the CPU-bound optimisation overhead. The optimisation technique performs an almost ...
Using Intelligent Software Agents to Query Heterogeneous Distributed Statistical Databases
The Internet and World-wide Web provide an ever-increasing source of diverse information. Softwar... more The Internet and World-wide Web provide an ever-increasing source of diverse information. Software agents have been developed by the Artificial Intelligence community with the aim of providing flexible and pro-active tools for users to access and utilise such information. In particular, such Agent Technology provides an
A hypermedia model is presented for the management of extendible multimedia information envisaged... more A hypermedia model is presented for the management of extendible multimedia information envisaged for tourist applications. The information is modelled according to the object-oriented methodology and organised into aggregation and specialisation / generalisation hierarchies, containing data that can be further enriched by authorised users - typically travel agents. The end-user conceives a hypermedia network, in which visited nodes are traced and navigation paths can be constructed on the fly.
Researchers working with vast quantities of information in a geographically distributed manner ar... more Researchers working with vast quantities of information in a geographically distributed manner are often confronted with problems of finding relevant information as well as colleagues with related interests. The MEMOIR project aims at assisting this collaboration by applying agent technology to user trails and documents. MEMOIR is an open architecture based on the existing Web infrastructure; in contrast to the Web, we treat links and trails as first class objects. Agents mine users' trails and links and also perform resource discovery tasks such as searching the Web. This paper describes the design, communication mechanism and implementation of the MEMOIR agent system which is currently being trialed in three end-user organisations. 1 Introduction Finding people with similar interests can be as difficult as efficiently accessing documents with related information. This is especially true for research oriented, globally distributed organizations such as pharmaceutical compan...
A popular approach for organising and accessing information is the hypermedia approach. One of th... more A popular approach for organising and accessing information is the hypermedia approach. One of the features that is weak in current hypermedia systems is the semantics of the nodes and links. In principle nodes are classified into categories while links represent abstract relationships among concepts. In this paper we present the implementation of parts of an object oriented data model that represents hypermedia information networks integrating semantics and multimedia information. The implementation refers to the parts of the data model that represent fuzzy knowledge in hypermedia networks. Special emphasis is put to the fuzzy features of the relationships among concepts.
HILDE : A Generic Platform for Building Hypermedia Training Applications 1
HILDE is a generic platform that aims to support the development of a wide range of educational a... more HILDE is a generic platform that aims to support the development of a wide range of educational applications that use multimedia assets such as video, image and sound. It supports reusability and provides flexibility by integrating a number of tools to support the tasks and the activities of four kinds of roles, namely: the MetaAuthoring role, the Authoring role, the trainee role and the tutor role. The tasks of the MetaAuthoring role are concerned with the organization and storage of the knowledge of a scientific domain. The Authoring role tasks concern the creation of a specific application. The trainee role encapsulates the activities of an individual trainee who attends a session of lessons either locally or remotely while his/her progress is automatically monitored and guided with intelligent educational techniques. A supervisor optionally monitors the trainees' progress and advises them, if needed. HILDE can meet the requirements of many and different organizations, since ...
Establishment of a European digital upper atmosphere server - DIAS project
The goal of DIAS (European digital upper atmosphere server) is to develop a pan-European digital ... more The goal of DIAS (European digital upper atmosphere server) is to develop a pan-European digital data collection on the state of the upper atmosphere, based on the existing five different historical data collections and on the real-time information provided by all five operating European digital ionospheric stations (digisondes). The operation of such a distributed information server will improve access to
A comparison of optimal scaling and preconditioning
ACM SIGNUM Newsletter, 1979
In this paper we shall compare two methods proposed for the equilibration of ill-conditioned line... more In this paper we shall compare two methods proposed for the equilibration of ill-conditioned linear systems when the coefficient matrix is small, dense and positive definite. Many workers have discussed ways of overcoming the problem of ill-conditioning and the main theme of all the methods is to transform the given system to one where the coefficient matrix is better conditioned.
A note on the construction of test matrices with equal eigenvalues
ACM SIGNUM Newsletter, 1977
What use to compute accurately the inverse of a matrix or the solution of a differential equation... more What use to compute accurately the inverse of a matrix or the solution of a differential equation if the output routine mangles the accurate answer? No program is better than the output service which prints its results. This note reports a horrid malfunction of a particular output facility. This error is no minor matter of incorrect rounding: I.B.M. Fortran H. Extended, level 2.1 (dated January 1975) will sometimes print, under G format, ".0000" where "i.000" should have been printedl Floating point numbers slightly less than unity, say .FFFFFF16 , are printed as zero under G Format. The same Erouble appears for numbers slightly less than other integer powers of I0. A few instances are shown at the end, the same number is shown in machine representation, as printed by F format, and as printed by two G formats. It is common for a program to have trouble at boundary situations, ends of loops, etc. In some sense, exact powers of i0 are boundary situations for a binary to decimal conversion; the behavior of the program changes as these boundaries are crossed. Presumably the trouble reported here is of this sort.
Integrated multimedia object and application modelling based on events and scenarios
Proceedings. International Workshop on Multi-Media Database Management Systems, 1995
An important issue in multimedia information systems is the lack of an integrated model for compl... more An important issue in multimedia information systems is the lack of an integrated model for complete representation of the multimedia objects and applications. In this paper we present an object oriented data base model that aims at representation of multimedia objects and applications. The application representation is based on composition (spatial and temporal) modelling and scenario modelling. An important feature
Digital Object Abstraction Layer: A Middleware for Building Federated Digital Libraries
Federated digital libraries focus on the development of common services over multiple, heterogene... more Federated digital libraries focus on the development of common services over multiple, heterogeneous digital repositories. To offer advanced services for federated digital libraries, there is a need to provide an abstract representation of digital objects stored within a specific repository, maintaining their semantics. In this paper a middleware for the development of federated digital libraries is introduced, named Digital Object Abstraction Layer (DOAL). It aims at (a) providing a unified abstract representation of digital objects stored in heterogeneous Digital Repositories independently of the software and location and (b) facilitating the development of complex reusable digital library service components. Its integration within federated digital library architecture promotes the creation of large scale distributed digital libraries, providing the same functionality as in the case where all digital content was stored in a single digital library system.
The Dowdy and Foster paper, "Comparative Models of the File Assignment Problem," COMPUTING SURVEY... more The Dowdy and Foster paper, "Comparative Models of the File Assignment Problem," COMPUTING SURVEYS, 14, 2 (June 1982), 287-313, discussed the optimal distribution of files among storage nodes in a computer system. Kollias and Hatzopoulos write to describe some additional models and studies of the file assignment problem-Ed.
Using parallelism and pipeline for the optimisation of join queries
Lecture Notes in Computer Science, 1992
Abstract: In this study we present a technique for the parallel optimisation of join queries, tha... more Abstract: In this study we present a technique for the parallel optimisation of join queries, that uses the offered coarse-grain parallelism of the underlying architecture in order to reduce the CPU-bound optimisation overhead. The optimisation technique performs an almost ...
A Cost Model for the Estimation Query Execution Time in a Parallel Environment Supporting Pipeline
Computing and Informatics / Computers and Artificial Intelligence, 1996
We propose a model for the estimation of query execution time in an environment supporting bushy ... more We propose a model for the estimation of query execution time in an environment supporting bushy and pipelined parallelism. We consider a parallel architecture of processes having private main memories, accessing a shared secondary storage and communicating to each other via a network. For this environment, we compute the cost of query operators when processed in isolation and when in pipeline mode. WE use those formulae to incrementally compute the cost of a query execution plan from its components. Our cost model can be incorporated to any optimizer for parallel query processing that considers parallel and pipelined execution of the query operators.
Agents for Querying Distributed Statistical Databases Over the Internet
International Journal on Artificial Intelligence Tools, 2002
Distributed database techniques and the Internet provide producers of statistics with a means to ... more Distributed database techniques and the Internet provide producers of statistics with a means to publish their data and metadata widely and make them available to a variety of users. Data matching to a user query and data access as well as data harmonization are some of the problems that should be solved. Intelligence is required in various stages of query answering and data matching. Moreover, the breadth and distributed nature of the Internet urge for a distributed approach. Agents seem to be the means by which both intelligence and distributed processing can be achieved. This paper presents a distributed approach for answering queries on statistical data that exist over the Internet using a multi-agent framework.
Parallel optimization of large join queries with set operators and aggregates in a parallel environment supporting pipeline
IEEE Transactions on Knowledge and Data Engineering, 1996
We propose a parallel optimizer for queries containing a large number of joins, as well as set op... more We propose a parallel optimizer for queries containing a large number of joins, as well as set operators and aggregate functions. The platform of execution is a shared-disk multiprocessor machine supporting bushy parallelism and pipeline. Our model partitions the query into almost independent subtrees that can be optimized simultaneously and applies an enhanced variation of the iterative improvement technique on
The main issues under consideration when facing network and distributed computing stem from the i... more The main issues under consideration when facing network and distributed computing stem from the interconnection of various elements residing in different sites over a local or wide area network (LAN or WAN). Ways of making possible the communication between different processes being executed on different computers have to be developed, a practice that is widely known under the term interprocess communication (IPC). These communication mechanisms are nothing more than abstractions, mainly software, enabling processes to ...
There are two main objectives of the DIAS (European Digital Upper Atmosphere Server) project. Fir... more There are two main objectives of the DIAS (European Digital Upper Atmosphere Server) project. First, it establishes a pan-European repository of raw and derived digital data describing the state of ionospheric part of the upper atmosphere, which is capable of ingesting real-time information and maintaining historical data collections provided by most operating ionospheric stations in Europe. Second, the DIAS system produces and distributes, based on the raw data collection, several operational products required by various user groups for ionospheric nowcasting, prediction and forecasting purposes. The project completed on May 2006 and the DIAS server operates since then continuously. The basic products that are delivered are real-time and historical ionograms from all DIAS-affiliated ionospheric stations, frequency plots and maps of the ionosphere over Europe based on the foF2, M(3000)F2, MUF and electron density parameters, as well as long term prediction and short term forecasting...
Parallel Optimisation of Join Queries Using a Technique of Exhaustive Nature 1
Abstract. In this study we present a technique for the parallel optimisation of join queries that... more Abstract. In this study we present a technique for the parallel optimisation of join queries that uses the offered coarse-grain parallelism of the underlying architecture in order to reduce the CPU-bound optimisation overhead. The optimisation technique performs an almost ...
Using Intelligent Software Agents to Query Heterogeneous Distributed Statistical Databases
The Internet and World-wide Web provide an ever-increasing source of diverse information. Softwar... more The Internet and World-wide Web provide an ever-increasing source of diverse information. Software agents have been developed by the Artificial Intelligence community with the aim of providing flexible and pro-active tools for users to access and utilise such information. In particular, such Agent Technology provides an
A hypermedia model is presented for the management of extendible multimedia information envisaged... more A hypermedia model is presented for the management of extendible multimedia information envisaged for tourist applications. The information is modelled according to the object-oriented methodology and organised into aggregation and specialisation / generalisation hierarchies, containing data that can be further enriched by authorised users - typically travel agents. The end-user conceives a hypermedia network, in which visited nodes are traced and navigation paths can be constructed on the fly.
Researchers working with vast quantities of information in a geographically distributed manner ar... more Researchers working with vast quantities of information in a geographically distributed manner are often confronted with problems of finding relevant information as well as colleagues with related interests. The MEMOIR project aims at assisting this collaboration by applying agent technology to user trails and documents. MEMOIR is an open architecture based on the existing Web infrastructure; in contrast to the Web, we treat links and trails as first class objects. Agents mine users' trails and links and also perform resource discovery tasks such as searching the Web. This paper describes the design, communication mechanism and implementation of the MEMOIR agent system which is currently being trialed in three end-user organisations. 1 Introduction Finding people with similar interests can be as difficult as efficiently accessing documents with related information. This is especially true for research oriented, globally distributed organizations such as pharmaceutical compan...
A popular approach for organising and accessing information is the hypermedia approach. One of th... more A popular approach for organising and accessing information is the hypermedia approach. One of the features that is weak in current hypermedia systems is the semantics of the nodes and links. In principle nodes are classified into categories while links represent abstract relationships among concepts. In this paper we present the implementation of parts of an object oriented data model that represents hypermedia information networks integrating semantics and multimedia information. The implementation refers to the parts of the data model that represent fuzzy knowledge in hypermedia networks. Special emphasis is put to the fuzzy features of the relationships among concepts.
HILDE : A Generic Platform for Building Hypermedia Training Applications 1
HILDE is a generic platform that aims to support the development of a wide range of educational a... more HILDE is a generic platform that aims to support the development of a wide range of educational applications that use multimedia assets such as video, image and sound. It supports reusability and provides flexibility by integrating a number of tools to support the tasks and the activities of four kinds of roles, namely: the MetaAuthoring role, the Authoring role, the trainee role and the tutor role. The tasks of the MetaAuthoring role are concerned with the organization and storage of the knowledge of a scientific domain. The Authoring role tasks concern the creation of a specific application. The trainee role encapsulates the activities of an individual trainee who attends a session of lessons either locally or remotely while his/her progress is automatically monitored and guided with intelligent educational techniques. A supervisor optionally monitors the trainees' progress and advises them, if needed. HILDE can meet the requirements of many and different organizations, since ...
Establishment of a European digital upper atmosphere server - DIAS project
The goal of DIAS (European digital upper atmosphere server) is to develop a pan-European digital ... more The goal of DIAS (European digital upper atmosphere server) is to develop a pan-European digital data collection on the state of the upper atmosphere, based on the existing five different historical data collections and on the real-time information provided by all five operating European digital ionospheric stations (digisondes). The operation of such a distributed information server will improve access to
A comparison of optimal scaling and preconditioning
ACM SIGNUM Newsletter, 1979
In this paper we shall compare two methods proposed for the equilibration of ill-conditioned line... more In this paper we shall compare two methods proposed for the equilibration of ill-conditioned linear systems when the coefficient matrix is small, dense and positive definite. Many workers have discussed ways of overcoming the problem of ill-conditioning and the main theme of all the methods is to transform the given system to one where the coefficient matrix is better conditioned.
A note on the construction of test matrices with equal eigenvalues
ACM SIGNUM Newsletter, 1977
What use to compute accurately the inverse of a matrix or the solution of a differential equation... more What use to compute accurately the inverse of a matrix or the solution of a differential equation if the output routine mangles the accurate answer? No program is better than the output service which prints its results. This note reports a horrid malfunction of a particular output facility. This error is no minor matter of incorrect rounding: I.B.M. Fortran H. Extended, level 2.1 (dated January 1975) will sometimes print, under G format, ".0000" where "i.000" should have been printedl Floating point numbers slightly less than unity, say .FFFFFF16 , are printed as zero under G Format. The same Erouble appears for numbers slightly less than other integer powers of I0. A few instances are shown at the end, the same number is shown in machine representation, as printed by F format, and as printed by two G formats. It is common for a program to have trouble at boundary situations, ends of loops, etc. In some sense, exact powers of i0 are boundary situations for a binary to decimal conversion; the behavior of the program changes as these boundaries are crossed. Presumably the trouble reported here is of this sort.
Integrated multimedia object and application modelling based on events and scenarios
Proceedings. International Workshop on Multi-Media Database Management Systems, 1995
An important issue in multimedia information systems is the lack of an integrated model for compl... more An important issue in multimedia information systems is the lack of an integrated model for complete representation of the multimedia objects and applications. In this paper we present an object oriented data base model that aims at representation of multimedia objects and applications. The application representation is based on composition (spatial and temporal) modelling and scenario modelling. An important feature
Digital Object Abstraction Layer: A Middleware for Building Federated Digital Libraries
Federated digital libraries focus on the development of common services over multiple, heterogene... more Federated digital libraries focus on the development of common services over multiple, heterogeneous digital repositories. To offer advanced services for federated digital libraries, there is a need to provide an abstract representation of digital objects stored within a specific repository, maintaining their semantics. In this paper a middleware for the development of federated digital libraries is introduced, named Digital Object Abstraction Layer (DOAL). It aims at (a) providing a unified abstract representation of digital objects stored in heterogeneous Digital Repositories independently of the software and location and (b) facilitating the development of complex reusable digital library service components. Its integration within federated digital library architecture promotes the creation of large scale distributed digital libraries, providing the same functionality as in the case where all digital content was stored in a single digital library system.
Hypermedia [1] is a fast growing field of research interest because
of the new aspects introduce... more Hypermedia [1] is a fast growing field of research interest because
of the new aspects introduced as far as it concerns the interaction
between the machine and the user. The innovative features mainly
refer to
- the presence of various information types (multimedia environment)
- the mapping of the semantic connections between information items
into a flexible, efficient representation.
This feature enables navigation through the information network in
a meaningful way. Our effort concentrated on the definition and partial
implementation of a flexible, general-purpose model through which
the full capabilities of object-oriented programming can be exploited.
Uploads
Papers by M. Hatzopoulos