Chiraz Trabelsi

Scalable mining of frequent tri-concepts from folksonomies

Mining frequent tri-concepts from folksonomies is an interesting problem with broad applications. Most of the previous tri-concepts mining based algorithms avoided a straightforward handling of the triadic contexts and paid attention to an unfruitful projection of the induced search space into dyadic contexts. As a such projection is very computationally expensive since several tri-concepts are computed redundantly, scalable mining of folksonomies remains a challenging problem. In this paper, we introduce a new algorithm, called Tricons, that directly tackles the triadic form of folksonomies towards a scalable extraction of tri-concepts. The main thrust of the introduced algorithm stands in the application of an appropriate closure operator that splits the search space into equivalence classes for the the localization of tri-minimal generators. These tri-minimal generators make the computation of the tri-concepts less arduous than do the pioneering approches of the literature.The experimental results show that the Tricons enables the scalable frequent tri-concepts mining over two real-life folksonomies .

Bridging Folksonomies and Domain Ontologies: Getting Out Non-taxonomic Relations

Social book marking tools are rapidly emerging on the Web as it can be witnessed by the overwhelming number of participants. In such spaces, users annotate resources by means of any keyword or tag that they find relevant, giving raise to lightweight conceptual structures \emph{aka} folksonomies. In this respect, needless to mention that ontologies can be of benefit for enhancing information retrieval metrics. In this paper, we introduce a novel approach for ontology learning from a \textit{folksonomy}, which provide shared vocabularies and semantic relations between tags. The main thrust of the introduced approach stands in putting the focus on the discovery of \textit{non-taxonomic} relationships. The latter are often neglected, even though they are of paramount importance from a semantic point of view. The discovery process heavily relies on triadic concepts to discover and select related tags and to extract and label non-taxonomically relationships between related tags and external sources for tags filtering and non-taxonomic relationships extraction. In addition, we also discuss a new approach to evaluate obtained relations in an automatic way against WordNet repository and presents promising results for a real world \textit{folksonomy}.

RssE-Miner: a new approach for efficient events mining from social media RSS feeds

Most of the new social media sites such as Twitter and Flickr are using RSS Feeds for sharing a wide variety of current and future real-world events. Indeed, RSS Feeds is considered as a powerful realtime means for real-world events sharing within the social Web. Thus, by identifying these events and their associated user-contributed social media resources, we can greatly improve event browsing and searching. However, a thriving challenge of events mining processes is owed to an efficient as well as a timely identification of events. In this paper, we are mainly dealing with event mining from heterogenous social media RSS Feeds. Therefore, we introduce a new approach, called RssE-Miner, in order to get out these events. The main thrust of the introduced approach stands in presenting a better trade-off between event mining accuracy and swiftness. Specifically, we adopted the probabilistic Naive Bayesian model within the exploitation of the rich context associated with social media Rss Feeds contents, including user-provided annotations (e.g., title, tags) and the automatically generated information (e.g., time) for efficiently mining future events. Carried out experiments over two real-world datasets emphasize the relevance of our proposal.

Exponential Ergodicity of the Jump-Diffusion CIR Process

In this paper we study the jump-diffusion CIR process (shorted as JCIR), which is an extension of the classical CIR model. The jumps of the JCIR are introduced with the help of a pure-jump Levy process \((J_t, t \ge 0)\). Under some suitable conditions on the Levy measure of \((J_t, t \ge 0)\), we derive a lower bound for the transition densities of the JCIR process. We also find some sufficient conditions under which the function \(V(x)=x\), \(x\ge 0\), is a Forster-Lyapunov function for the JCIR process. This allows us to prove that the JCIR process is exponentially ergodic.

Positive Harris recurrence and exponential ergodicity of the basic affine jump-diffusion

Abstract In this article, we find the transition densities of the basic affine jump-diffusion (BAJD), which has been introduced by Duffie and Gârleanu as an extension of the CIR model with jumps. We prove the positive Harris recurrence and exponential ergodicity of the BAJD. Furthermore, we prove that the unique invariant probability measure π of the BAJD is absolutely continuous with respect to the Lebesgue measure and we also derive a closed-form formula for the density function of π.

BGRT : une nouvelle base générique de règles d'association triadiques : Application à l'autocomplétion de requêtes dans les folksonomies

Le tagging social s’est recemment impose dans le paysage du web collaboratif (Web 2.0) comme un support a l’organisation des ressources partagees, permettant aux utilisateurs de categoriser leurs ressources en leurs associant des mots clefs, appeles tags. La structure ainsi creee, baptisee sous le nom de folksonomie, est assimilee a un hypergraphe triparti d’utilisateurs, de tags et de ressources. Dans ce papier, nous exploitons ces triplets pour introduire une nouvelle definition d’une base generique de regles d’association triadiques, appelee . Nous montrons que l’utilisation de ces regles generiques pour l’autocompletion de requetes permet de mettre en exergue la pertinence des folksonomies et leur interet reel pour la recherche d’information. Les premiers resultats obtenus sur une folksonomie reelle s’averent prometteurs et ouvrent de nombreuses perspectives.

Community Detection in Multi-relational Bibliographic Networks

In this paper, we introduce a community detection approach from heterogeneous multi-relational network which incorporate the multiple types of objects and relationships, derived from a bibliographic networks. The proposed approach performs firstly by constructing the relation context family (RCF) to represent the different objects and relations in the multi-relational bibliographic networks using the Relational Concept Analysis (RCA) methods; and secondly by exploring such RCF for community detection. Experiments performed on a dataset of academic publications from the Computer Science domain enhance the effectiveness of our proposal and open promising issues.

CoMRing: A Framework for Community Detection Based on Multi-relational Querying Exploration☆

Abstract Community detection in multi-relational bibliographic networks is an important issue. There has been a surge of interest in community detection focusing on analyzing the linkage or topological structure of these networks. However, communities identified by these proposed approaches, commonly reflect the strength of connections between networks nodes and neglect considering the interesting topics or the venues, i.e., conferences or journals, shared by these community members, i.e, authors. To tackle this drawback, we present in this paper a new approach called CoMRing for community detection from heterogeneous multi-relational network which incorporate the multiple types of objects and relationships, derived from a bibliographic networks. We firstly propose to construct the Concept Lattice Family (CLF) to model the different objects and relations in the multi-relational bibliographic networks using the Relational Concept Analysis (RCA) methods. Then after we introduce a new method, called QueryExploration, that explores such CLF for community detection. Carried out experiments on real-datasets enhance the effectiveness of our proposal and open promising issues.

A Probabilistic Approach for Events Identification from Social Media RSS Feeds

Social Media RSS feeds are the most up-to-date and inclusive releases of information on current events used by the new social media sites such as Twitter and Flickr. Indeed, RSS feeds are considered as a powerful realtime means for real-world events sharing within the social Web. By identifying these events and their associated social media resources, we can greatly improve event browsing and searching. However, a thriving challenge of events identification from such releases is owed to an efficient as well as a timely identification of events. In this paper, we are mainly dealing with event identification from heterogenous social media RSS feeds. In this respect, we introduce a new approach in order to get out these events. The main thrust of the introduced approach stands in achieving a better tradeoff between event identification accuracy and swiftness. Specifically, we adopted the probabilistic Naive Bayes model within the exploitation of stemming and feature selection techniques. Carried out experiments over two real-world datasets emphasize the relevance of our proposal and open many issues.

Hypergraph fuzzy minimals transversals mining: A new approach for social media recommendation

User preference discovery aims to detect the patterns of user preferences for various topics of interest or items such as movie genre or category. Preferences discovery is a crucial stage in the development of intelligent personalization systems. Although a variety of studies have been proposed in the literature addressing a wide range of applications such as recommender systems or personalized search, only a few of them have considered the management of imprecision in the representation of user and item features. This paper aims to address the above issue by using fuzzy sets. The paper proposes a general framework for preferences discovery through fuzzy sets and fuzzy models and it introduces a new algorithm for representing and discovering fuzzy user interest profile. Based on the results of the empirical evaluation, the proposed approach outperforms two well-known recommendation approaches in terms of well-known quality assessment metrics, namely: discounted cumulative gain, precision, recall, as well as F1-measure.