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.

Collaborative tagging is a simple and effective method for organizing and sharing web resources using human created metadata. It has arisen out of the need for an efficient method of personal organization, as the number of digital resources in everyday lives increases. While tagging has become a proven organization scheme through its popularity and widespread use on the Web, little is known about its implications and how it may effectively be applied in different situations. This is due to the fact that tagging has evolved through several iterations of use on social software websites, rather than through a scientific or an engineering design process. The research presented in this thesis, through investigations in the domain of e-learning, seeks to understand more about the scientific nature of collaborative tagging through a number of human subject studies. While broad in scope, touching on issues in human computer interaction, knowledge representation, Web system architecture, e-learning, metadata, and information visualization, this thesis focuses on how collaborative tagging can supplement the growing metadata requirements of e-learning. I conclude by looking at how the findings may be used in future research, through using information based in the emergent social networks of social software, to automatically adapt to the needs of individual users.

Collaborative tagging is a simple and effective method for organizing and sharing web resources using human created metadata. It has arisen out of the need for an efficient method of personal organization, as the number of digital resources in everyday lives increases. While tagging has become a proven organization scheme through its popularity and widespread use on the Web, little is known about its implications and how it may effectively be applied in different situations. This is due to the fact that tagging has evolved through several iterations of use on social software websites, rather than through a scientific or an engineering design process. The research presented in this thesis, through investigations in the domain of e-learning, seeks to understand more about the scientific nature of collaborative tagging through a number of human subject studies. While broad in scope, touching on issues in human computer interaction, knowledge representation, Web system architecture, e-learning, metadata, and information visualization, this thesis focuses on how collaborative tagging can supplement the growing metadata requirements of e-learning. I conclude by looking at how the findings may be used in future research, through using information based in the emergent social networks of social software, to automatically adapt to the needs of individual users.

Every single day, lots of users actively participate in social media sites (e.g., Facebook, YouTube, Last.fm, Flicker, etc.) upload photos, videos, share bookmarks, write blogs and annotate/comment on content provided by others. With the recent proliferation of social media sites, users are overwhelmed by the huge amount of available content. Therefore, organizing and retrieving appropriate multimedia content is becoming an increasingly important and challenging task. This challenging task led a number of research communities to concentrate on social tagging systems (also known as folksonomy) that allow users to freely annotate their media items (e.g., music, images, or video) with any sort of arbitrary words, referred to as tags. Tags assist users to organize their own content, as well as to find relevant content shared by other users. In this thesis, we first analyze how useful a folksonomy is for improving personalized services such as tag recommendation, tag-based search and item annotation. We then propose two new algorithms for social media retrieval and tag recommendation respectively. The first algorithm computes the latent preferences of tags for users from other similar tags, as well as latent annotations of tags for items from other similar items. We then seamlessly map the tags onto items, depending on an individual user’s query, to find the most desirable content relevant to the user’s needs. The second algorithm improves tag-recommendation and item annotation by adapting the Katz measure, a path-ensemble based proximity measure, for the use in social tagging systems. In this algorithm we model folksonomy as a weighted, undirected tripartite graph. We then apply the Katz measure to this graph, and exploit it to provide personalized tag recommendation for individual users. We evaluate our algorithms on two real-world folksonomies collected from Last.fm and CiteULike. The experimental results demonstrate that the proposed algorithms improve the search and the recommendation performance, and obtain significant gains in cold start situations where relatively little information is known about a user or an item

Every single day, lots of users actively participate in social media sites (e.g., Facebook, YouTube, Last.fm, Flicker, etc.) upload photos, videos, share bookmarks, write blogs and annotate/comment on content provided by others. With the recent proliferation of social media sites, users are overwhelmed by the huge amount of available content. Therefore, organizing and retrieving appropriate multimedia content is becoming an increasingly important and challenging task. This challenging task led a number of research communities to concentrate on social tagging systems (also known as folksonomy) that allow users to freely annotate their media items (e.g., music, images, or video) with any sort of arbitrary words, referred to as tags. Tags assist users to organize their own content, as well as to find relevant content shared by other users. In this thesis, we first analyze how useful a folksonomy is for improving personalized services such as tag recommendation, tag-based search and item annotation. We then propose two new algorithms for social media retrieval and tag recommendation respectively. The first algorithm computes the latent preferences of tags for users from other similar tags, as well as latent annotations of tags for items from other similar items. We then seamlessly map the tags onto items, depending on an individual user’s query, to find the most desirable content relevant to the user’s needs. The second algorithm improves tag-recommendation and item annotation by adapting the Katz measure, a path-ensemble based proximity measure, for the use in social tagging systems. In this algorithm we model folksonomy as a weighted, undirected tripartite graph. We then apply the Katz measure to this graph, and exploit it to provide personalized tag recommendation for individual users. We evaluate our algorithms on two real-world folksonomies collected from Last.fm and CiteULike. The experimental results demonstrate that the proposed algorithms improve the search and the recommendation performance, and obtain significant gains in cold start situations where relatively little information is known about a user or an item

L’évolution du Web et de ses applications subit depuis quelques années une mutation vers les technologies qui incluent la dimension sociale comme entité de première classe. Nous témoignons dans le passage du Web 1.0 au Web 2.0 puis au Web 3.0, 4.0 etc.. que les utilisateurs et les réseaux sociaux qui se forment sont au centre de cette évolution. Le web exhibe aussi toutes les caractéristiques d’un système complexe. Ces propriétés systèmes complexes et cette dimension sociale doivent être prises en considération lors de la conception et le développement des applications web. Considérons le cas des systèmes de tagging ou d’étiquetage collaboratifs. Ces systèmes sont un exemple de systèmes complexes, auto-organisés et socialement conscients. Le paradigme des systèmes multi-agents coordonné par les mécanismes d’auto-organisations a été utilisé d’une façon effective pour la conception et modélisation des systèmes complexes. Les systèmes de tagging collaboratifs actuels ne prennent pas l’avantage complet de leurs caractéristiques systèmes complexes, surtout dans l’adaptation à leur environnement et l’émergence de nouvelles fonctionnalités. Dans ce travail de thèse, nous proposons un modèle pour la conception et développement d’un nouveau système d’étiquetage collaboratif MySURF (My Similar Users, Resources, Folksonomies), utilisant une approche multi-agents gouvernée par la coévolution des organisations sociales et spatiales des agents. Nous montrons comment ce système proposé offre plusieurs nouvelles fonctionnalités qui peuvent améliorer les systèmes d’étiquetage collaboratifs actuels. / The evolution of the Web and its applications has undergone in the last few years a mutation towards technologies that include the social dimension as a first class entity. We are witnessing in the evolution of the web from the web 1.0 to web 2.0 to web 3.0 and eventually web 4.0 that the users, their interactions and the emerging social networks are in the center of this evolution. The web also exhibits all the characteristics of a complex system. These complex systems properties and this social dimension must be taken into consideration in the design and the development of new web applications. Let us consider the case of collaborative tagging systems. These systems are an example of complex, self-organized and socially aware systems. The multi-agent systems paradigm coordinated by self-organizations mechanisms was used in an effective way for the design and modeling of the complex systems. Current collaborative tagging systems do not take full advantage of the characteristics of complex systems, especially in adapting to their environment and the emergence of new features. In this thesis, we propose a model for the design and development of a new collaborative tagging system MySURF (My Similar Users, Resources, Folksonomies), using a multi-agent system approach governed by the coevolution of the social and spatial organization of the agents. We show how the proposed system offers several new features that can improve current collaborative tagging systems.

Made available in DSpace on 2016-06-02T19:07:09Z (GMT). No. of bitstreams: 1 AMARAL_Anderson_2014.pdf: 3835882 bytes, checksum: 2fdc6c0cc705277a06fbcce246dd4b79 (MD5) Previous issue date: 2014-12-03 / Aspects of social tagging have been studied in the last years aiming at addressing concepts, applicability and techniques related to the generation and management of tags as key elements for the resource research and classification. Social tagging has been recognized as an important alternative for the description of resources available on the Web. Concerning on the rapid growth of resources and their information on the Web, it has become increasingly difficult to find and organize such massive data by traditional methods as based on directories classification. The process of social tagging has been investigated to improve the cataloging of resources, being recommendation algorithms of tags, visualization techniques and automatic generation of tags, among others have been explored for this purpose. Within the e-learning context, the social tagging may assist on building of metadata of learning objects. Thus, all the advantages linked to this model of resources organization, has emerged as solutions for e-learning systems. The main contribuition of a teaching-learning environment is the creation, storing and maintenance of the catalog of learning resources that can be checked and updated continuously. This work proposes a tags recommendation system, improving the process of cataloguing of learning resources on the web. Three experiments were carried out with technician students of a informatics course to validate the proposal: the first one examined the social tagging activity without using the tag recommendation system; the second one was conducted to evaluate the tagging activity, but including the tags recommendation system; and the third one verified the students navigation from the tagged resources created from the first and the second experiment, in order to check that the support of the recommendation was effective in the descriptor vocabulary of the resource. / Aspectos da marca¸c ao social t em sido estudados nos ´ultimos anos com o objetivo de abordar conceitos, aplicabilidade e t´ecnicas relacionadas `a gera¸c ao e gest ao de tags como elementos-chave para a pesquisa de recursos e classifica¸c ao. A marca¸c ao social tem sido reconhecida como uma importante alternativa para a descri¸c ao dos recursos dispon´ıveis na Web. No que diz respeito sobre o r´apido crescimento dos recursos e suas informa¸c oes na Web, tornou-se cada vez mais dif´ıcil de encontrar e organizar essa grande massa de dados por m´etodos tradicionais com base na classifica¸c ao em diret´orios. O processo de marca¸c ao social tem sido investigado para melhorar a cataloga¸c ao de recursos, sendo que estrat´egias como algoritmos de recomenda¸c ao de tags, t´ecnicas de visualiza¸c ao e gera¸c ao autom´atica de tags, entre outras foram exploradas para esta finalidade. Dentro do contexto de e-learning, a marca¸c ao social pode ajudar na constru¸c ao de metadados de objetos de aprendizagem. Assim, todas as vantagens associadas a este modelo de organiza¸c ao de recursos, surgiu como solu¸c ao para sistemas de e-learning. A principal contribui¸c ao de um ambiente de ensino-aprendizagem ´e a cria¸c ao, armazenamento e manuten¸c ao do cat´alogo de recursos de aprendizagem que pode ser verificado e atualizado continuamente. Este trabalho prop oe um sistema de recomenda¸c ao de tags, melhorando o processo de cataloga¸c ao de recursos de aprendizagem na Web. Tr es experimentos foram realizados com os alunos de um curso t´ecnico de inform´atica para validar a proposta: o primeiro examinou a atividade de marca¸c ao social sem usar o sistema de recomenda¸c ao de tags; o segundo foi realizado para avaliar a atividade de marca¸c ao, mas incluindo o sistema de recomenda¸c ao de tags; e o terceiro verificou a navega¸c ao dos estudantes a partir dos recursos marcados e criados a partir do primeiro e do segundo experimentos, a fim de verificar que o apoio da recomenda¸c ao foi eficaz no vocabul´ario descritor dos recursos.

Suite à l’avènement du Web 2.0, le rôle de l’internaute s’est vu modifier, passant de consommateur passif à acteur à part entière. De nouvelles fonctionnalités ont vu le jour augmentant considérablement les possibilités d’interaction avec le système. Parmi celles-ci, le collaborative tagging permet à l’utilisateur de décrire l’information en ligne par l’attribution de mots-clés (ou tags), la particularité étant que ces tags ne sont pas uniquement accessibles aux tagueurs eux-mêmes mais à l’ensemble des internautes. L’octroi de tags à une ressource lui offre donc de multiples chemins d’accès exploitables par la communauté internet tout entière. Régulièrement comparé à l’indexation « professionnelle », le collaborative tagging soulève une question essentielle :cette nouvelle pratique contribue-t-elle favorablement à la description et, par extension, à la recherche d’informations sur internet ?<p>Tous les types d’informations ne pouvant être étudiés, la présente dissertation se focalise sur l’information médicale scientifique utilisée par les médecins dans le cadre de leur pratique professionnelle. Elle propose, dans un premier temps, de mesurer le potentiel des tags assignés dans deux systèmes de collaborative tagging (Delicious et CiteULike) à décrire l’information en les comparant à des descripteurs attribués par des professionnels de l’information pour un même échantillon de ressources. La comparaison a mis en lumière l’exploitabilité des tags en termes de dispositifs de recherche d’informations mais a néanmoins révélé des faiblesses indéniables par rapport à une indexation réalisée par des professionnels à l’aide d’un langage contrôlé.<p>Dans un second temps, la dissertation s’est intéressée aux utilisateurs finaux en quête d’informations, c’est-à-dire les médecins, afin de déterminer dans quelle mesure un système de collaborative tagging (CiteULike) peut assister ces derniers lors de leur recherche d’informations scientifiques. Pour ce faire, des entretiens individuels combinant interview semi-structurée et expérimentation ont été organisés avec une vingtaine de médecins. Ils ont fourni des indications riches et variées quant à leur adoption effective ou potentielle d’un système de collaborative tagging dans le cadre de leurs pratiques informationnelles courantes.<p>Enfin, cette dissertation se propose d’aller au-delà de l’étude des tags et du phénomène de collaborative tagging dans son ensemble. Elle s’intéresse également aux compétences informationnelles des médecins observés en vue d’alimenter la réflexion sur les formations qui leur sont dispensées tout au long de leurs études mais également durant leur parcours professionnel. / Doctorat en Information et communication / info:eu-repo/semantics/nonPublished