We present a case study of establishing a description infrastructure for media processing systems. The description infrastructure consists of an internal metadata model and access tools for using it. Based on an analysis of requirements, we selected, out of a set of candidates, MPEG-7 as the basis of our metadata model.

The developments of computer systems in terms of memory and capacity in the last years, as well as the overwhelming phenomena of wide band networks, have revived in the international scientific community the drive to investigate and test novel audiovisual content analysis tools and algorithms. The principal aim of these techniques is to provide users with powerful automatic classification and description tools allowing for considerable cost savings and efficiency in a wide range of content search and retrieval processes. Quite obviously, broadcasters and content providers represent a prime target for the results of these efforts due to their relevant investments and interests in audiovisual archives. This paper proposes a critical review of the state of the art in content analysis from the point of view of the applicability in real contexts and operating environments. In addition it presents some results carried on at RAI in the field of automatic documentation and classification of archive content based on these tools.

Rich News, a system that augments news broadcasts with textual content, is described. The system identifies individual stories in news broadcasts, and annotates them with related content from the World Wide Web.../...

This paper describes the part of the PrestoSpace project dedicated to the study and development of the Metadata Access and Delivery (MAD) system. The mission of the MAD system, inside the wider perspective of the PrestoSpace factory, is to generate, validate and deliver to the archive users metadata created through the employment of both automatic and manual information extraction tools. Automatic tools include audiovisual content analysis and semantic analysis of text extracted by automatic speech recognition (ASR). The MAD publication platform provides access and search facilities to the imported and newly produced metadata in a synergic and easy-to-use interface.

The purpose of this article is to provide EBU Members with basic information on the possible benefits of employing content analysis techniques for documenting their television and radio archives. It is based on the considerable experience gained by RAI over recent years in this field and in the PrestoSpace project. The article also looks analytically at the impact these new archive documentation techniques will have on traditional working practices.

Manual camera motion annotation has been performed by three groups using a tool provided by Joanneum Research. Some types of content made it difficult or impossible for human annotators to describe the camera motion. The comparison of annotations of the same content done by two groups shows significant differences for some features. We discuss the questions that arise from the results of the manual annotation.../...

In recent years tree kernels have been proposed for the automatic learning of natural language applications. Unfortunately, they show (a) an inherent super linear complexity and (b) a lower accuracy than traditional attribute/value methods. In this paper, we show that tree kernels are very helpful in the processing of natural language as (a) we provide a simple algorithm to compute tree kernels in linear average running time and (b) our study on the classification properties of diverse tree kernels show that kernel combinations always improve the traditional methods.../...

Work done by JRS & Partners in the PrestoSpace Restoration Work Area.

In this paper, a system, RitroveRAI, addressing the general problem of enriching a multimedia news stream with semantic metadata is presented. News metadata here are explicitly derived from transcribed sentences or implicitly expressed into a topical category automatically detected. The enrichment process is accomplished by searching the same news expressed by different agencies reachable over the Web.../...

The Rich News system, that can automatically annotate radio and television news with the aid of resources retrieved from the World Wide Web, is described. Automatic speech recognition gives a temporally precise but conceptually inaccurate annotation model. Information extraction from related web news sites gives the opposite: conceptual accuracy but no temporal data. Our approach combines the two for temporally accurate conceptual semantic annotation of broadcast news.../...

Web services tend to offer functional and non-functional requirements and capabilities in an agreed, machine-readable format. The target is to provide automated services for discovering, selection and binding of information as a native capability of middleware and applications. However, the major limitations are due to the lack of clear and processable semantics. In this paper, a system, RitroveRAI, addressing the general problem of enriching a multimedia news stream with semantic metadata is presented.../...

Web services tend to offer functional and non-functional requirements and capabilities in an agreed, machine-readable format. The target is to provide automated services for discovering, selection and binding of information as a native capability of middleware and applications. However, the major limitations are due to the lack of clear and processable semantics. In this paper, a system, RitroveRAI, addressing the general problem of enriching a multimedia news stream with semantic metadata is presented.../...

The Rich News system for semantically annotating television news broadcasts and augmenting them with additional web content is described. Online news sources were mined for material reporting the same stories as those found in television broadcasts, and the text of these pages was semantically annotated using the KIM knowledge management platform.../...

MPEG-7 is an excellent choice for the description of audiovisual content due to its flexibility and comprehensiveness. The drawback is that these properties also increase the complexity of descriptions and cause ambiguities which hinder interoperability. In order to partly solve these problems, profiles and levels have been proposed, but the definitions of the adopted profiles lack semantic constraints which are necessary for interoperability. We propose a profile for detailed description of audiovisual content that can be used in a broad range of applications.../...

Recent work on the design of automatic systems for semantic role labeling has shown that feature engineering is a complex task from a modeling and implementation point of view. Tree kernels alleviate such complexity as kernel functions generate features automatically and require less software development for data extraction. In this paper, we study several tree kernel approaches for both boundary detection and argument classification.../...

An international metadata conference in association with The Bridgeman Art Library, BAPLA and MDA. Hosted by The MILE Project, this conference workshops and showcases the latest cataloguing standards for image archiving, with key speakers including Murtha Baca, Head of Vocabulary Program and Digital Resource Management at The Getty Institute and author of Cataloguing Cultural Objects, and Monika Hagedorn-Saupe, Deputy Director of Stiftung Preussischer Kulturbesitz. The basis of this event is the need to address the key issues associated with the global harmonisation of metadata, and to tackle these via a combination of focussed debate and workshops on key standards such as CDWA Lite, museumdat XML and OAI. We believe it is especially important that with the current increasing interest in metadata we should focus attention on metadata for opening access to the major cultural resources represented by national collections.