Search results

Multimedia retrieval is a complex and to some extent still unexplored area. Based on a full year of e‐mail requests addressed to a large film archive this study analyses what types of information needs real users have and how these needs are expressed. The findings include that the requesters make use of a broad range of need attributes in specifying their information needs. These attributes relate to the production, content, subject, context and screening of films. However, a few attributes – especially title, production year and director – account for the majority of the attribute instances. Further, as much as 43 per cent of the requests contain no information about the context that gives rise to the request. The current indexing of the archived material is restricted to production‐related attributes, and access to the material is, thus, frequently dependent on the archivists’ extensive knowledge of the archived material and films in general.

Affordable mobile devices with video playback functionality are rapidly growing in the market. Current wireless and third generation communication networks enable smoother and higher quality streaming video. With the support of these technologies, most participants in telecom value‐added service chain are planning to shift their business focus to a more profitable and appealing area, mobile TV. Previous work that survey on users' behavior while consuming mobile TV has indicated that users normally watch brief and casual contents, and not the full program. However, most of the current services adopt a “push” approach since users passively receive pre‐defined contents, rather than pulling the interesting topics and segments. In order to promote a more enjoyable and rewarding watching experience, this paper will propose a framework to support a fully interactive mobile TV. The main goal is to enable users to: 1) visually locate interesting topics across multiple genres (such as news, sports and entertainment) and 2) fully control the playback flow of the multimedia items while selecting the most interesting segments. A web‐based system has been developed to implement and test the effectiveness of the proposed framework in a wireless and mobile setting.

The aim of this paper is to examine related studies to identify which retrieval methods are supported by current digital cultural heritage libraries. In this way it is hoped to provide a direction for future cultural heritage applications to provide more complete and/or improved retrieval functionality.

The methodology of this paper is based on introducing the general concept of image‐based retrieval systems as well as their retrieval methods. Then, users' needs are discussed to illustrate the demands of semantic‐based retrieval. After the retrieval methods have been presented, current digital cultural heritage libraries are examined in terms of their supported retrieval methods that allow users to query images.

Current digital cultural heritage libraries mostly provide only general retrieval methods based on image‐based low‐level features, i.e. query by image contents. Very few consider other retrieval methods such as browsing and semantic‐based retrieval. In addition, none of the current systems provide all possible retrieval methods for users.

This study is the first one to examine image‐based retrieval methods in digital cultural heritage libraries. This study supports the improvement of retrieval functionality for digital cultural heritage libraries in the future.

The purpose of this paper is to introduce an interpreting schema and semantic description framework for a collection of images of Xilankapu, a traditional Chinese form of embroidered fabric and brocade artwork.

First, the authors interpret the artwork of Xilankapu through Gillian Rose’s “four site” theory by presenting how the brocades were made, how the patterns of Xilankapu are classified and the geometrical abstraction of visual images. To further describe the images of this type of brocade, this paper presents semantic descriptions that include objective–non-objective relations and a multi-layered semantic framework. Furthermore, the authors developed corresponding methods for scanning, storage and indexing images for retrieval.

As exploratory research on describing, preserving and indexing images of Xilankapu in the context of the preservation of cultural heritage, the authors collected 1,000+ images of traditional Xilankapu, classifying and storing some of the images in a database. They developed an index schema that combines concept- and content-based approaches according to the proposed semantic description framework. They found that the framework can describe, store and preserve semantic and non-semantic information of the same image. They relate the findings of this paper to future research directions for the digital preservation of traditional cultural heritages.

The framework has been designed especially for brocade, and it needs to be extended to other types of cultural image.

The semantic description framework can describe connotative semantic information on Xilankapu. It can also assist the later information retrieval work in organizing implicit information about culturally related visual materials.

Along with the ever growingWeb is the proliferation of objectionable content, such as sex, violence, racism, etc. We need efficient tools for classifying and filtering undesirable web content. In this paper, we investigate this problem through WebGuard, our automatic machine learning based pornographic website classification and filtering system. Facing the Internet more and more visual and multimedia as exemplified by pornographic websites, we focus here our attention on the use of skin color related visual content based analysis along with textual and structural content based analysis for improving pornographic website filtering. While the most commercial filtering products on the marketplace are mainly based on textual content‐based analysis such as indicative keywords detection or manually collected black list checking, the originality of our work resides on the addition of structural and visual content‐based analysis to the classical textual content‐based analysis along with several major‐data mining techniques for learning and classifying. Experimented on a testbed of 400 websites including 200 adult sites and 200 non pornographic ones, WebGuard, our Web filtering engine scored a 96.1% classification accuracy rate when only textual and structural content based analysis are used, and 97.4% classification accuracy rate when skin color related visual content based analysis is driven in addition. Further experiments on a black list of 12 311 adult websites manually collected and classified by the French Ministry of Education showed that WebGuard scored 87.82% classification accuracy rate when using only textual and structural content‐based analysis, and 95.62% classification accuracy rate when the visual content‐based analysis is driven in addition. The basic framework of WebGuard can apply to other categorization problems of websites which combine, as most of them do today, textual and visual content.

[Editor's note: this contribution is a presentation given by the author at the International Federation for Information and Documentation Education and Training Pre‐Conference held in Jaipur, India, 12–13 October, 1998. The complete paper in the FID/ET Proceedings was co‐authored by Dr. S.P. Sood and Smt. Prabha Krishnan. The matrix showing the knowledge base and the required computer and information literacy skills is recommended for use as a checklist by all institutions involved in education for library and information science].

Large amount of data are stored in image format. Image retrieval from bulk databases has become a hot research topic. An alternative method for efficient image retrieval is proposed based on a combination of texture and colour information. The main purpose of this paper is to propose a new content based image retrieval approach using combination of color and texture information in spatial and transform domains jointly.

Various methods are provided for image retrieval, which try to extract the image contents based on texture, colour and shape. The proposed image retrieval method extracts global and local texture and colour information in two spatial and frequency domains. In this way, image is filtered by Gaussian filter, then co-occurrence matrices are made in different directions and the statistical features are extracted. The purpose of this phase is to extract noise-resistant local textures. Then the quantised histogram is produced to extract global colour information in the spatial domain. Also, Gabor filter banks are used to extract local texture features in the frequency domain. After concatenating the extracted features and using the normalised Euclidean criterion, retrieval is performed.

The performance of the proposed method is evaluated based on the precision, recall and run time measures on the Simplicity database. It is compared with many efficient methods of this field. The comparison results showed that the proposed method provides higher precision than many existing methods.

The comparison results showed that the proposed method provides higher precision than many existing methods. Rotation invariant, scale invariant and low sensitivity to noise are some advantages of the proposed method. The run time of the proposed method is within the usual time frame of algorithms in this domain, which indicates that the proposed method can be used online.

This paper surveys theoretical and practical issues associated with a particular type of information retrieval problem, namely that where the information need is pictorial. The paper is contextualised by the notion of a visually stimulated society, in which the ease of record creation and transmission in the visual medium is contrasted with the difficulty of gaining effective subject access to the world's stores of such records. The technological developments which, in casting the visual image in electronic form, have contributed so significantly to its availability are reviewed briefly, as a prelude to the main thrust of the paper. Concentrating on still and moving pictorial forms of the visual image, the paper dwells on issues related to the subject indexing of pictorial material and discusses four models of pictorial information retrieval corresponding with permutations of the verbal and visual modes for the representation of picture content and of information need.

The overwhelming speed and scale of digital media production greatly outpace conventional indexing methods by humans. The management of Big Data for e-library speech resources requires an automated metadata solution. The paper aims to discuss these issues.

A conceptual model called semantic ontologies for multimedia indexing (SOMI) allows for assembly of the speech objects, encapsulation of semantic associations between phonic units and the definition of indexing techniques designed to invoke and maximize the semantic ontologies for indexing. A literature review and architectural overview are followed by evaluation techniques and a conclusion.

This approach is only possible because of recent innovations in automated speech recognition. The introduction of semantic keyword spotting allows for indexing models that disambiguate and prioritize meaning using probability algorithms within a word confusion network. By the use of AI error-training procedures, optimization is sought for each index item.

Validation and implementation of this approach within the field of digital libraries still remain under development, but rapid developments in technology and research show rich conceptual promise for automated speech indexing.

The SOMI process has been preliminarily tested, showing that hybrid semantic-ontological approaches produce better accuracy than semantic automation alone.

Even as testing proceeds on recorded conference talks at the University of Tebessa (Algeria), other digital archives can look toward similar indexing. This will mean greater access to sound file metadata.

Huge masses of spoken data, unmanageable for a human indexer, can prospectively find semantically sorted and prioritized indexing – not transcription, but generated metadata – automatically, quickly and accurately.

This research aims to identify users' goals and strategies when searching for podcasts and their impact on the design of podcast retrieval technology. In particular, the paper seeks to explore the potential to address user goals with indexing based on podcast metadata and automatic speech recognition (ASR) transcripts.

The paper conducted a user study to obtain an overview of podcast search behaviour and goals, using a multi‐method approach of an online survey, a diary study, and contextual interviews. In a subsequent podcast retrieval experiment, the paper investigated the retrieval performance of the two choices of indexing features for search goals identified during the study.

The paper found that study participants used a variety of search strategies, partially influenced by available tools and their perceptions of these tools. Furthermore the experimental results revealed that retrieval using ASR transcripts performed significantly better than metadata‐based searching. However, a detailed result analysis suggested that the efficacy of the indexing methods was search‐goal dependent.

The research constitutes a step towards a future framework for investigating user needs and addressing them in an experimental set‐up. It was primarily qualitative and exploratory in nature.

Podcast search engines require evidence about suitable indexing methods in order to make an informed decision concerning whether it is worth the resources to generate speech recognition transcripts.

Systematic studies of podcast searching have not previously been reported. Investigations of this kind hold the potential to optimise podcast retrieval in the long term.