Manel Mezghani

A user profile modelling using social annotations: a survey

As social networks are growing in terms of the number of users, resources and interactions; the user may be lost or unable to find useful information. Social elements could avoid this disorientation like the social annotations (tags) which become more and more popular and contribute to avoid the disorientation of the user. Representing a user based on these social annotations has showed their utility in reflecting an accurate user profile which could be used for a recommendation purpose. In this paper, we give a state of the art of characteristics of social user and techniques which model and update a tag-based profile. We show how to treat social annotations and the utility of modelling tag-based profiles for recommendation purposes.

Information Quality in Social Networks: A Collaborative Method for Detecting Spam Tweets in Trending Topics.

In Twitter based applications such as tweet summarization, the existence of ill-intentioned users so-called spammers imposes challenges to maintain high performance level in those applications. Conventional social spammer/spam detection methods require significant and unavoidable processing time, extending to months for treating large collections of tweets. Moreover, these methods are completely dependent on supervised learning approach to produce classification models, raising the need for ground truth data-set. In this paper, we design an unsupervised language model based method that performs collaboration with other social networks to detect spam tweets in large-scale topics (e.g. hashtags). We experiment our method on filtering more than 6 million tweets posted in 100 trending topics where Facebook social network is accounted in the collaboration. Experiments demonstrate highly competitive efficiency in regards to processing time and classification performance, compared to conventional spam tweet detection methods.

Evaluating Seed Selection for Information Diffusion in Mobile Social Networks

The integration of social networks with mobile communication has led to the rise of a new paradigm, the mobile social network (MSN). Recently, MSN has emerged as a new hot spot of research attracting much interest from both academia and industrial sectors. For instance, MSN opens new horizon for information diffusion-based applications such as viral marketing. Thus, it is a fundamental issue to select an efficient subset of seed-nodes (i.e. initial sources) in a MSN such that targeting them initially will maximize the information diffusion to interested nodes. This paper studies the problem of identifying the best seeds through whom the information can be diffused in the network in order to maximize the content utility (i.e. a quantitative metric that determines how satisfied are the users). A multi- layer model that combines the social relationships and the mobile network in order to design an efficient information diffusion is proposed. Based on this multi-layer model, different seed selection approaches are proposed for information diffusion environment (e.g. mobile advertising) where users have heterogeneous interests for the different information generated in the network. Simulation results show the effectiveness of multi-layer based seed selection approaches comparing to a classical approach.

Towards Generating Spam Queries for Retrieving Spam Accounts in Large-Scale Twitter Data

Twitter, as a top microblogging site, has became a valuable source of up-to-date and real-time information for a wide range of social-based researches and applications. Intuitively, the main factor of having an acceptable performance in those recherches and applications is the working and relying on information having an adequate quality. However, given the painful truth that Twitter has turned out a fertile environment for publishing noisy information in different forms. Consequently, maintaining the condition of high quality is a serious challenge, requiring great efforts from Twitter’s administrators and researchers to address the information quality issues. Social spam is a common type of the noisy information, which is created and circulated by ill-intentioned users, so-called social spammers. More precisely, they misuse all possible services provided by Twitter to propagate their spam content, leading to have a large information pollution flowing in Twitter’s network. As Twitter’s anti-spam mechanism is not both effective and immune towards the spam problem, enormous recherches have been dedicated to develop methods that detect and filter out spam accounts and tweets. However, these methods are not scalable when handling large-scale Twitter data. Indeed, as a mandatory step, the need for an additional information from Twitter’s servers, limited to a few number of requests per 15 min time window, is the main barrier for making these methods too effective, requiring months to handle large-scale Twitter data. Instead of inspecting every account existing in a given large-scale Twitter data in a sequential or randomly fashion, in this paper, we explore the applicability of information retrieval (IR) concept to retrieve a sub-set of accounts having high probability of being spam ones. Specifically, we introduce a design of an unsupervised method that partially processes a large-scale of tweets to generate spam queries related to account’s attributes. Then, the spam queries are issued to retrieve and rank the highly potential spam accounts existing in the given large-scale Twitter accounts. Our experimental evaluation shows the efficiency of generating spam queries from different attributes to retrieve spam accounts in terms of precision, recall, and normalized discounted cumulative gain at different ranks.

De l'influence de l'enrichissement de profil utilisateur sur la propagation de buzz dans les médias sociaux. Expérimentations sur Delicious

L’utilisateur est la source principale de l’information diffusee dans les medias sociaux et mais il est, en meme temps, influence par les informations partagees sur les reseaux. Le phenomene de buzz, c’est-a-dire faire « du bruit » autour d’une information (fait ou rumeur) pour que plusieurs utilisateurs soient interesses par celle-ci simultanement, peut etre defini comme une information populaire dans un temps specifique. Notre etude porte sur l’influence de l’enrichissement dynamique de profil utilisateur (Mezghani et al., 2014) sur la propagation des buzz avec application au reseau social Delicious. Delicious contient des annotations sociales (tags) fournies par les utilisateurs et qui contribuent a influencer les autres utilisateurs afin de suivre certaines informations ou de les utiliser. Notre etude suit la methodologie suivante : 1) nous analysons la propagation des tags consideres comme des buzz dans le temps, 2) nous appliquons l’enrichissement dynamique de profil utilisateur et nous etudions l’influence de cet enrichissement dans la propagation de buzz, 3) nous analysons si l’approche d’enrichissement anticipe la propagation de buzz. Nous montrons dans cet article l’interet, lors de l’enrichissement de profil, de filtrer les informations afin de proposer des resultats pertinents a l’utilisateur et eviter de « mauvaises » recommandations.

Analyse du comportement d'annotation du réseau social d'un utilisateur pour la détection des intérêts - Application sur Delicious

L’utilisateur social est caracterise par son activite sociale comme le partage d’informations et l’etablissement de relations avec d’autres utilisateurs. Avec l’evolution du contenu social, l’utilisateur a besoin d’informations plus precises qui refletent ses interets. Nous nous concentrons sur la detection des interets de l’utilisateur qui sont des elements cles pour ameliorer l’adaptation (recommandation, personnalisation, etc.). L’originalite de notre approche est basee sur la proposition d’une nouvelle technique de detection des interets qui analyse le reseau des relations d’un utilisateur et aussi la precision de leurs comportements d’annotation dans le but de selectionner les tags qui refletent reellement le contenu des ressources. L’approche proposee a ete testee et evaluee sur la base de donnees sociales Delicious. Pour l’evaluation, nous comparons le resultat issu de notre approche utilisant le comportement d’annotation des personnes proches (le reseau egocentrique ou les communautes) avec les informations connues de l’utilisateur (son profil). Une evaluation comparative avec une approche classique (basee sur les tags) de detection des interets montre que l’approche proposee fournit de meilleurs resultats.