An Analysis of XML Database Solutions for the Management
Of MPEG-7 Media Descriptions
MOHSIN ARSALAN
SZABIST
KARACHI, PAKISTAN

Abstract: MPEG-7 based applications will be set up in the near future due to its promising standard for the description of multimedia contents. Therefore management of large amount of mpeg-7 complaint media description is what is needed. Mpeg documents are XML documents with media description schemes defined in xml schema. In this paper are mentioned some of the solutions that help make a management system for mpeg-7. Furthermore these solutions are compared and analyzed. The result shows how limited today’s databases are, when comes the case of mpeg-7.
Keywords: MPEG-7, XML, SCHEMA
1. INTRODUCTION The paper is organized in the following manner that sec1 is the introduction to what mpeg-7 is. Then the solutions for its management are mentioned. Then the solutions are analyzed and finally in the end the conclusion is mentioned. 1.1. MPEG-7: Mpeg (moving picture expert group) is the creator of the well known mpeg1 and mpeg2 and mpeg4. Mpeg-7 is an ISO standard which is developed by mpeg group. The mpeg-7 standard formally named “multimedia content description interface” provides a rich of standardized tools which describes a multimedia content. Humans as well as automatic systems process audiovisual information which is within the mpeg7 scope. Audiovisual description tools (the metadata elements and their structure and relationships that are defined by the standard in the form of descriptors and description schemes) to create descriptions (i.e., a set of instantiated Description Schemes and their corresponding Descriptors at the users will), which will form the basis for applications enabling the needed effective and efficient access (search, filtering and browsing) to multimedia content. Mpeg-7 with these description schemes gives an extensive description of multimedia contents on a technical as well as semantic level. 1.2. MPEG-7 media descriptions and description schemes: (REFER TO FIGURE [1]) Mpeg7 DDL can be considered as another general purpose schema definition language for xml as it is a superset of xml schema. In the mpeg 7 DDL schema definitions the description of media scheme is defined by complex type (described in xml schema). The complex type of mpeg7 is that they can be derived from within another.

FIGURE [1]: Overview of predefined MPEG-7 media description schemes.

1. Observations: The two main observations related to the management of mpeg7 media descriptions are: * Mpeg 7 does not support a fixed number of media description schemes due to its ready to use media description schemes. * Mpeg7 scheme and schema definition of xml are similar and does not define an independent xml conceptual model. This similarity between xml representation and conceptual model managing mpeg7 media descriptions causes managing mpeg7 media descriptions effectively means managing xml documents. Reverse engineering such a model is a tedious job.

2. Representation of media descriptions: The properties and characteristics of mpeg7 should be kept in mind when proposing solutions for its effective management. Fine-grained representations for its media description and for basic contents typed representation are proposed solutions. Versioning is also required for some applications so as to represent media descriptions in them.

3.3. Fine grained representation: Fine grained representation is important to manage mpeg7 media descriptions in xml database solution. Such media descriptions of fine granularity require such a structure. This model represents the hierarchy of the media description’s elements and attributes values. This in return allows applications to access desired parts efficiently. The drawback to it is that effort must be spent for the decomposition of media descriptions when they are imported as well as exported from the database. Other structures like DOM etc have also been proposed. 3.4. Typed representation: Large portions in the mpeg7 media description are non textual. For those basic parts typed representation is required so as to represent them in an xml database. The drawback in order to represent nontextual data in textual form is that it consumes more space and that it also is less efficient. It also burdens applications because of constant and explicit conversion of accessing non textual contents. This constant work can also over head performance and pile up inconsiderably. 3.5. Versioning: The phases of content production, post production, consumption and delivery are a repetitive process in order to produce multimedia content. During this repetitive process the contents are constantly rearranged and edited. If media descriptions are with the production process along with different versions of the content this is where versioning comes. In short
“A versioning-aware representation allows applications to comfortably access all current and past versions of a media description, to trace the predecessor and successor relationships between the different versions of a description, and to derive new versions of a media description from existing ones” [1]
The drawback is that existing xml databases do not offer versions. 3. Summary:
The drawbacks of the proposed solutions are that the nontextual data must be read by applications as they are instead of text. Matrix, numeric, vectors etc; which make up technical metadata; must be accessed as they are and not text. Another issue is that value indexing support offered by database solutions is not sufficient enough. Also keeping the facts in mind that fine grained concurrency control, fine grained access control along with versioning are also lacking. Native xml databases might be good when access to mpeg7 data is concerned but is weak when extensibility with new functionality and index structures are kept in consideration
5. CONCLUSION In this paper the requirements to manage such data in an xml document is mentioned along with what today’s xml database offers and how extensibility actually allows current databases to support such data. The deficiencies of current databases allow systems to not only validate mpeg media descriptions against their description schemes but also to perform more sophisticated access optimizations. For better management we therefore need a better management system.. a need for a new generation of xml database solution is required. These databases must be able to recognize the importance of schema type information exploitation. But text, paths, index structures must not be forgotten along with classic d.b.m.s functionality.

References: [1]: “An Analysis of XML Database Solutions for the Management of MPEG-7 Media Descriptions”
UTZ WESTERMANN AND WOLFGANG KLAS
University of Vienna
A.C.M