The use of multimedia content on the web has grown significantly in recent years. Websites such as Facebook, YouTube and Flickr cater for enormous amounts of multimedia content uploaded by users. This vast amount of multimedia content requires comprehensive content modelling otherwise retrieving relevant content will be challenging. Modelling multimedia content can be an extremely time consuming task that may seem impossible particularly when undertaken by individual users. However, the advent of Web 2.0 and associated communities, such as YouTube and Flickr, has shown that users appear to be more willing to collaborate in order to take on enormous tasks such as multimedia content modelling. Harnessing the power of communities to achieve comprehensive content modelling is the primary focus of this research. The aim of this thesis is to explore collaborative multimedia content modelling and in particular the effectiveness of existing multimedia content modelling tools, taking into account the key development challenges of existing collaborative content modelling research and the associated modelling tools. Four research objectives are pursued in order to achieve this; first, design a user experiment to study users’ tagging behaviour with existing multimedia tagging tools and identify any relationships between such user behaviour; second, design and develop a framework for MPEG-7 content modelling communities based on the results of the experiment; third, implement an online service as a proof of concept of the framework; fourth, validate the framework through the online service during a repeat of the initial user experiment. This research contributes first, a conceptual model of user behaviour visualised as a fuzzy cognitive map and, second, an MPEG-7 framework for multimedia content modelling communities (MC2) and its proof of concept as an online service. The fuzzy cognitive model embodies relationships between user tagging behaviour and context and provides an understanding of user priorities in the description of content features and the relationships that exist between them. The MC2 framework, developed based on the fuzzy cognitive model, is deep-rooted in user content modelling behaviour and content preferences. A proof of concept of the MC2 framework is implemented as an online service in which all metadata is modelled using MPEG-7. The online service is validated, first, empirically with the same group of users and through the same experiment that led to the development of the fuzzy cognitive model and, second, functionally against the folksonomy and MPEG-7 content modelling tools used in the initial experiment. The validation demonstrates that MC2 has the advantages without the shortcomings of existing multimedia tagging tools by harnessing the ease of use of folksonomy tools while producing comprehensive structured metadata.

Libraries, private and public, offer valuable resources to library patrons; however, formulating library queries to retrieve relevant results can be difficult. This occurs because when using a library catalog for library searches, patrons often do not know the exact keywords to be included in a query that match the rigid subject terms (chosen by the Library of Congress) or terms in other fields of a desired library catalog record. These improperly formulated queries often translate into a high percentage of failed searches that retrieve irrelevant results or no results at all. This explains why frustrated library patrons nowadays rely on Web search engines to perform their searches first, and upon obtaining the initial information, such as titles, subject areas, or authors, they query the library catalog. This searching strategy is an evidence of failure of today's library systems. In solving this problem, we propose an enhanced library system, called EnLibS, which allows partial, similarity matching of (i) tags defined by ordinary users at a folksonomy site which describe the content of books and (ii) keywords in a library query to improve the searches on library catalogs. The proposed library system allows patrons to post a query Q with commonly-used words and ranks the retrieved results according to their degrees of resemblance with Q. Experimental results show that EnLibS (i) reduces the amount of queries that retrieve no results, (ii) obtains high precision in retrieving and accuracy in ranking relevant results, and (iii) achieves a processing time comparable to existing library catalog search engines.

No description available.

The simplicity and flexibility of tagging allows users to collaboratively create large, loosely structured repositories of Web resources. One of its main drawbacks is the need for manual formulation of tags for each posted resource. This task can be eased by a tag recommendation system, the objective of which is to propose a set of tags for a given resource, user pair. Tag recommendation is an interesting and well-defined practical problem. Its main features are constant interaction with users and availability of large amounts of tagged data. Given the opportunities (e.g., rich user feedback) and limitations (e.g., real-time response) of the tag recommendation setting, we defined six requirements for a practically useful tag recommendation system. We present a conceptual design and system architecture of a hybrid tag recommendation system, which meets all these requirements. The system utilizes the strengths of various tag sources (e.g., resource content and user profiles) and the relations between concepts captured in tag co-occurrence graphs mined from collaborative actions of users. The architecture of the proposed system is based on a text indexing engine, which allows the system to deal with large datasets in real time, while constantly adapting its models to newly added posts. The effectiveness and efficiency of the system was evaluated for six datasets representing a broad range of collaborative tagging systems. The experiments confirmed the high quality of results and practical usability of the system. In a comparative study the system outperformed a state-of-the-art algorithm based on tensor factorization for the most representative datasets applicable to both methods. The experiments on the characteristics of tagging data and the performance of the system allowed us to find answers to important research questions adapted from the general area of recommender systems. We confirmed the importance of infrequently used tags in the recommendation process and proposed solutions to overcome the cold start problem in tag recommendation. We demonstrated that a parameter tuning approach makes a hybrid tag recommendation system adaptable to various datasets. We also revealed the importance of the utilization of a feedback loop in the tag recommendation process.

This research is based on the assumption that social media sites – such as Facebook, Twitter and YouTube – have changed the way in which memes (culturally transmitted ideas) are transmitted or spread. Qualitative research is used in this study, particularly a literature study that applies and weighs the theories against a purposive sample of selected film examples between 2009 and 2010, to answer the question: what is the relationship between social media sites and the spread of memes? To answer this question, the way in which social media aids the spread of memes is analysed. It is argued that social media has provided the most fertile environment for the replication of memes to date. Also analysed is the way in which social media sites are represented in films, as well as the ways in which the offline lives of the characters are affected by their social media profiles, specifically as depicted in films Julie & Julia (Ephron 2009), Trust (Schwimmer 2010), and Catfish (Joost & Schulman 2010). Lastly, it is argued that one can construct an entirely new persona, or merely experiment with different aspects of who you are, on social media sites. This online persona can be influenced by the memes one chooses to spread online. Naturally, all memes carry connotations, values and judgements. These memes collaborate with one’s profile and, thus, the connotations attached to the memes one shares are then associated with one’s online persona (profile). This research is deemed relevant, also in a South African context, because the spread of memes does not respect national or international boundaries. / Dissertation (MA)--University of Pretoria, 2014. / lk2014 / Visual Arts / MA / Unrestricted

Le système de tags pour un système d’organisation des connaissances centralise et fournit les tags qui peuvent être utilisés pour classer, partager et rechercher des connaissances sur le web pour l’utilisation personnelle ou organisationnelle. Bien que les études précédentes aient pensé à améliorer le système de tags visuels en utilisant des icônes, il existe dans ce cas le problème de reconnaissance, de mémorisation et de désorientation. Notre recherche se consacre à la recherche d'une nouvelle approche pour améliorer la représentation des tags et surtout de leur structure, dans un système où les icônes bien structurées pourront améliorer l'efficacité de tagage en considérant la qualité et la rapidité. Ce système de tags iconiques s’organise sur un LVD (Langage Visuel Distinctif) lui-même basé sur le modèle Hypertopic pour la représentation de cartes de thèmes multipoints de vue développé par l’équipe Tech-CICO. Cette solution est proposée pour améliorer principalement l'interprétation sémiotique du sens de l’icone et renforcer la compréhension et l’usage de la structure de tags dans un système informatisé de partage des connaissances, notamment pour gérer et partager les tags iconiques sur une plate-forme collaborative / Tags systems for Knowledge Organization System centralize and provide the tags that can be employed in classifying, sharing and seeking knowledge on the web for personal or organizational use. However, an increased variety of vocabularies and languages cause connections between tags and documents marked by textual tags to become less and less distinctive, making the use and reuse of tags systems even harder. Although previous attempts have been made onto visual tags system by using icons, it caused the disorientation when users facing with plant of isolated symbols. Our research dedicates to searching a new approach to improve the representation of tags and their structure in a tags system, where well-structured icons enhance the tagging effectiveness by considering tagging quality and tagging speed. The LVD (Visual Distinctive Language)-based iconic tags system is proposed and presented in this thesis to bring amelioration mainly from semiotic interpretation of tag meaning and graphical code of tag structure. The arrangement of icons is as well another interesting topic that was deal with in our research to offers a more complete definition of iconic tags system. Apart from modeling and evaluating the LVD-based iconic tags system we have considered the way to build up such icon system in today’s cooperative knowledge sharing context and made it possible to manage and share iconic tags on a collaborative plate-form

Cette recherche propose une analyse des usages des systèmes de tagging sur le Web. Au delà d'un simple outil personnel de « catégorisation ordinaire » des ressources en ligne, les classifications produites à partir de l'indexation libre de tags, nommées folksonomies, permettent aux internautes de produire des « prises » physiques et interprétatives quideviennent un support aux formes de navigation sociale. Ces réseaux de traces numériques permettent aux internautes de se repérer et d'agir dans l'univers complexe et abondant que représente le Web, mais elles sont également un support de coordination avec d'autres internautes. Les tags apparaissent comme des appuis conventionnels qui permettent decoordonner les actions au sein de collectifs à géométrie variable, plus ou moins étendus et identifiés, dont les membres partagent des centres d’intérêts et un vocabulaire commun. A partir d'une approche socio-technique nous étudions le couplage entre conception et usage de ces dispositifs, et montrons que le tagging propose une voie alternative, encore enconstruction, aux politiques existantes d'architecture et d’accessibilité des ressources du Web / This research proposes an anlysis of tagging systems uses on the Web. Tagging is not a simple personnal tool of web ressources categorization but classifications emerging from free tags indexation, named folksonomies, allow Internet users to create physical and interpretative supports which produce new forms of social navigation. Those tagging networks give the opportunity to Internet users to navigate and to act in the complex and abundant Web universe, but they are also coordination supports with other Internet users. Tags appear as conventionnal supports which allow coordination of actions in variable-geometry collectives. In those more or less identified and large collectives, members share common vocabulary and interests. Based on a socio-technical approach, we study coupling between conception and uses of tagging systems. We explain that tagging proposes an under construction and alternative way to architecture and accessibility politics of the Web ressources

Radio frequency identification (RFID) is a technology used in many applications for the identification of objects. This thesis presents a concept of how libraries could use RFID technology to locate physical items within the library. The ability to locate items within the library changes the way users interact with physical material, creates new ways of user collaboration, and influences the ability to browse the shelves for physical items. Several implementation scenarios are presented in detail and implications on collaboration and browsing are analyzed.

Radio frequency identification (RFID) is a technology used in many applications for the identification of objects. This thesis presents a concept of how libraries could use RFID technology to locate physical items within the library. The ability to locate items within the library changes the way users interact with physical material, creates new ways of user collaboration, and influences the ability to browse the shelves for physical items. Several implementation scenarios are presented in detail and implications on collaboration and browsing are analyzed.

Orientador: André Santanchè / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Computação / Made available in DSpace on 2018-08-20T14:39:12Z (GMT). No. of bitstreams: 1 Alves_HugoAugusto_M.pdf: 1442382 bytes, checksum: 258a15cb04d919b837754948bad5829e (MD5) Previous issue date: 2012 / Resumo: Um número crescente de repositórios na web se baseia em metadados na forma de rótulos (tags) para organizar e classificar o seu conteúdo. Os usuários destes sistemas associam livremente tags a recursos do sistema - e.g., URLs, imagens, marcadores. O termo folksonomia se refere a esta classificação coletiva, que emerge do processo de rotulação (tagging) realizado por usuários interagindo em ambientes sociais na web. Uma das maiores qualidades das folksonomias é a sua simplicidade de uso pela ausência de um vocabulário controlado. Folksonomias crescem de forma orgânica, refletindo o conhecimento da comunidade de usuários. Por outro lado, esta falta de estrutura leva a dificuldades em operações de organização e descoberta de conteúdo. Melhores resultados podem ser obtidos se forem consideradas as relações semânticas entre os rótulos. Por esta razão, vários trabalhos foram propostos com o objetivo de relacionar ontologias e folksonomias, combinando a estrutura sistematizada das ontologias à semântica latente das folksonomias. Enquanto em uma direção algumas abordagens criam "ontologias sociais" a partir dos dados das folksonomias, em outra direção algumas abordagens conectam rótulos a ontologias preexistentes. Em ambos os casos nota-se uma unidirecionalidade, ou seja, um modelo apenas dá suporte ao enriquecimento do outro. Nossa proposta, por outro lado, é bidirecional. Ontologias e folksonomias são fundidas em uma nova entidade, que chamamos de "ontologia folksonomizada", combinando aspectos complementares de ambas. O conhecimento formal e projetado das ontologias é fundido com a semântica latente dos dados sociais. Nesta dissertação apresentamos nossa ontologia folksonomizada e seus desdobramentos. Nós introduzimos um framework formal para a análise de trabalhos relacionados, a fim de confrontá-los com a nossa abordagem. Além das melhorias nas operações de indexação e descoberta, que foram validadas em experimentos práticos, nós propomos uma técnica chamada 3E Steps para dar suporte à evolução de ontologias usando dados de folksonomias. Nós também implementamos o protótipo de uma ferramenta para a construção de ontologias folksonomizadas e para dar suporte à revisão de ontologias / Abstract: An increasing number of web repositories relies on tag-based metadata to organize and classify their content. The users of these systems freely associate tags with resources of the system - e.g., URLs, images, and bookmarks. The term folksonomy refers to this collective classification, which emerges from tagging carried by users interacting in web social environments. One of the major strengths of folksonomies is their simplicity due to the absence of a controlled vocabulary. Folksonomies grow organically, reflecting the knowledge of a community of users. On the other hand, this lack of structure leads to difficulties in operations of content organization and discovery. Better results can be obtained if we take into account the semantic relations among tags. For this reason, many proposals were developed aiming to relate ontologies and folksonomies, combining the systematized structure of ontologies to the latent semantics of folksonomies. While in one direction some approaches build "social ontologies" from folksonomic data, in the other direction some approaches connect tags to existing ontologies. In both cases they are unidirectional approaches, i.e., one model is used only to support the enrichment of the other. Our proposal, on the other hand, is bidirectional. Ontologies and folksonomies are fused in a new entity, we call "folksonomized ontology", which combines complementary aspects of both. The formal and engineered knowledge of ontologies is fused with the latent semantics of social data. In this dissertation we present our folksonomized ontology and its outcomes. We introduce here a formal framework to analyze the related work, confronting it with our approach. Besides the improvements in indexing and discovery operations, which are validated by practical experiments, we propose a 3E Steps technique to support ontology evolvement by using folksonomic data. We also have implemented a tool prototype to build folksonomized ontologies and to support ontology review / Mestrado / Ciência da Computação / Mestre em Ciência da Computação