ABSTRACTS OF PUBLICATIONS

Home: MIRROR.

Index: VLDB 99, TWLT-14, DS-8, SPIE 98, IDEAS 98, IRSG 98, Displays, CTIT Report No. 97-35, IDMS 96.

[VLDB99]

Arjen P. de Vries, Mark G.L.M. van Doorn, Henk M. Blanken, Peter M.G. Apers, The Mirror MMDBMS architecture, technical demo at VLDB 99 Edinborough, 1999.

Abstract

Handling large collections of digitized multimedia data, usually referred to as multimedia digital libraries, is a major challenge for information technology. The Mirror DBMS is a research database system that is developed to better understand the kind of data management that is required in the context of multimedia digital libraries (see also the mirror home page). Its main features are an integrated approach to both content management and (traditional) structured data management, and the implementation of an extensible object-oriented logical data model on a binary relational physical data model. The Mirror DBMS provides the primitives for managing the images and its meta data, as well as the probabilistic inference required during the interaction with the user. New feature models, different clustering algorithms, or different query modification techniques, can easily be added or modified. Hence, a variety of multimedia retrieval systems can be implemented by simply changing the sequence of Moa expressions issued in the Mirror DBMS.

[TWLT14]

Arjen P. de Vries, Mirror: Multimedia Query Processing in Extensible Databases, in 14th Twente Workshop on Language Technology. Language Technology in Multimedia Information Retrieval, Enschede, The Netherlands, December 1998.

Abstract

The Mirror project investigates the implications of multimedia information retrieval on database design. We assume a modern extensible database system with extensions for feature based search techniques. The multimedia query processor has to bridge the gap between the user's high level information need and the search techniques available in the database. We therefore propose an iterative query process using relevance feedback. The query processor identifies which of the available representations are most promising for answering the query. In addition, it combines evidence from different sources. Our multimedia retrieval model is a generalization of a well-known text retrieval model. We discuss our prototype implementation of this model, based on Bayesian reasoning over a concept space of automatically generated clusters. The experimentation platform uses structural object-orientation to model the data and its meta-data flexibly, without compromising efficiency and scalability. We illustrate our approach with some first experiments with text and music retrieval.

Keywords

Multimedia Information Retrieval, Digital Libraries, Multimedia Query Processing, Inference Network Retrieval Model

[DS8]

Arjen P. de Vries, Annita N. Wilschut, On the integration of IR and databases, accepted as short paper on 8th IFIP 2.6 Working Conference on Database Semantics (DS-8).

Abstract

Integration of information retrieval (IR) in database management systems (DBMSs) has proven difficult. Previous attempts to integration suffered from inherent performance problems, or lacked desirable separation between logical and physical data models. To overcome these problems, we discuss a database approach based on structural object-orientation. We implement IR techniques using extensions in an object algebra called MOA. MOA has been implemented on top of the database backend Monet, a state-of-the-art high-performance database kernel with a binary relational interface. Our prototype implementation of the inference network retrieval model using MOA and Monet demonstrates the feasibility of this approach. We conclude with a discussion of the advantages of our database design.

The original, longer draft version of the paper also discusses optimization in the context of extending an object algebra.

Keywords

retrieval, distributed object model.

[SPIE98]

Arjen P. de Vries, Henk M. Blanken, Database technology and the management of multimedia data in Mirror, in Multimedia Storage and Archiving Systems III, volume 3527 of Proceedings of SPIE, Boston, November 1998.

Abstract

Multimedia digital libraries require an open distributed architecture instead of a monolithic database system. In the Mirror project, we use the Monet extensible database kernel to manage different representations of multimedia objects. To maintain independence between content, meta-data, and the creation of meta-data, we allow distribution of data and operations using CORBA. This open architecture introduces new problems for data access. From an end user's perspective, the problem is how to search the available representations to fulfill an actual information need; the conceptual gap between human perceptual processes and the meta-data is too large. From a system's perspective, several representations of the data may semantically overlap or be irrelevant. We address these problems with an iterative query process and active user participation through relevance feedback. A retrieval model based on inference networks assists the user with query formulation. The integration of this model into the database design has two advantages. First, the user can query both the logical and the content structure of multimedia objects. Second, the use of different data models in the logical and the physical database design provides data independence and allows algebraic query optimization. We illustrate query processing with a music retrieval application.

Keywords

Digital libraries, extensible databases, distributed databases, database architecture, multimedia query processing, multimedia information

[IDEAS98]

Arjen P. de Vries, Brian Eberman, David E. Kovalcin, The design and implementation of an infrastructure for multimedia digital libraries, in Proceedings of the 1998 International Database Engineering & Applications Symposium, pages 103-110, Cardiff, UK, July 1998.

Abstract

We develop an infrastructure for managing, indexing and serving multimedia content in digital libraries. This infrastructure follows the model of the web, and thereby is distributed in nature. We discuss the design of the Librarian, the component that manages meta data about the content. The management of meta data has been separated from the media servers that manage the content itself. Also, the extraction of the meta data is largely independent of the Librarian. We introduce our extensible data model and the daemon paradigm that are the core pieces of this architecture. We evaluate our initial implementation using a relational database. We conclude with a discussion of the lessons we learned in building this system, and proposals for improving the flexibility, reliability, and performance of the system.

Keywords

digital libraries, multimedia databases, multimedia modeling, content-based

[IRSG98]

Arjen P. de Vries, Henk M. Blanken, The Relationship between IR and Multimedia Databases, Accepted for publication at IRSG98.

Abstract

Modern extensible database systems support multimedia data through ADTs. However, because of the problems with multimedia query formulation, this support is not sufficient. Multimedia querying requires an iterative search process involving many different representations of the objects in the database. The support that is needed is very similar to the processes in information retrieval. Based on this observation, we develop the miRRor architecture for multimedia query processing. We design a layered framework based on information retrieval techniques, to provide a usable query interface to the multimedia database. First, we introduce a concept layer to enable reasoning over low-level concepts in the database. Second, we add an evidential reasoning layer as an intermediate between the user and the concept layer. Third, we add the functionality to process the users' relevance feedback. We then adapt the inference network model from text retrieval to an evidential reasoning model for multimedia query processing. We conclude with an outline for implementation of miRRor on top of the Monet extensible database system.

[DISPLAYS98]

Arjen P. de Vries, Gerrit C. van der Veer, Henk M. Blanken, Let's talk about it: Dialogues with multimedia databases. Database support for human activity, Displays, 18(4):215-220, 1998.

Abstract

We describe two scenarios of user tasks in which access to multimedia data plays a significant role. Because current multimedia databases cannot support these tasks, we introduce three new requirements on multimedia databases: multimedia objects should be active objects, querying is an interaction process, and query processing uses multiple representations. We discuss three techniques to handle multimedia objects as active objects. Also, we introduce a promising database architecture to meet the new user requirements. Agents within the database handle objects' representations, and a search engine on top of a conventional database handles relevance feedback and multiple representations.

Keywords

multimedia databases, multimedia modeling, human computer interaction, content-based retrieval, relevance feedback.

[CTIT97]

Arjen P. de Vries, Intelligent Television: A testbed for multimedia information filtering, CTIT Technical Report series, No. 97-35.

Abstract

We set up an environment in which a database is filled with television data from several information channels. Philips Sound and Vision provides a prototype television that is very suited for this application. The combination of television and personal computer is used to investigate new paradigms to enhance the traditional television interface. Multimedia database technology provides the basis for integration of the program guide in the television interface and the automatical recording of potentially interesting programs. The television helps the user to cope with his information overload. The experimental implementation reveals improvements for the development of the intelligent television. Moreover, the system provides for a testbed environment for multimedia information filtering.

Keywords

multimedia, information filtering, multimedia databases, television, information overload.

[IDMS96]

Arjen P. de Vries, Television Information Filtering through Speech Recognition, Interactive Distributed Multimedia Systems and Services (IDMS '96), Berlin, Germany, 1996, pages 59-69.

Abstract

The problem of information overload can be solved by the application of information filtering to the huge amount of data. Information on radio and television can be filtered using speech recognition of the audio track. A prototype system using closed captions has been developed on top of the INQUERY information access system. The challange of integrating speech recognition and information retrieval into a working system is a big one. The open problems are the selection of a document representation model, the recognition and selection of indexing features for speech retrieval and dealing with the erroneous output of recognition processes.

Keywords

multimedia, multimedia representation, content-based retrieval, information filtering, automatic indexing, speech recognition, content analysis, probabilistic information retrieval.
Index: VLDB 99, TWLT-14, DS-8, SPIE 98, IDEAS 98, IRSG 98, Displays, CTIT Report No. 97-35, IDMS 96.

Home: MIRROR.

Last updated: $Id: abstracts.html,v 1.1 1998/02/25 10:09:01 arjen Exp arjen $
Maintained by: arjen@cs.utwente.nl