Rushed Kanawati

Université Sorbonne Paris Nord

We present a position paper about our concept for an artificial intelligence (AI) and data streaming platform for the agricultural sector. The goal of our project is to support agroecology in terms of carbon farming and biodiversity protection by providing an AI and data streaming platform called Gaia-AgStream that accelerates the adoption of AI in...

In this work, we explore applying a link prediction approach to tag recommendation in broad folksonomies. The original idea of the approach is to mine the dynamic of the tagging activity in order to compute the most suitable tag for a given user and a given resource. The tagging history of each user is modeled by a temporal sequence of bipartite gr...

This chapter presents the problem of link prediction in complex networks. It provides general description, formal definition of the problem and applications. It gives a state-of-art of various existing link prediction approaches concentrating more on topological approaches. It presents the main challenges of link prediction task in real networks. T...

In the field of web mining and web science, as well as data science and data mining there has been a lot of interest in the analysis of (social) networks. With the growing complexity of heterogeneous data, feature-rich networks have emerged as a powerful modeling approach: They capture data and knowledge at different scales from multiple heterogene...

Abstract The growing availability of multirelational data gives rise to an opportunity for novel characterization of complex real-world relations, supporting the proliferation of diverse network models such as Attributed Graphs, Heterogeneous Networks, Multilayer Networks, Temporal Networks, Location-aware Networks, Knowledge Networks, Probabilisti...

Graph clustering techniques are very useful for detecting densely connected groups in large graphs. Many existing graph clustering methods mainly focus on the topological structure, but ignore the vertex properties. Existing graph clustering methods have been recently extended to deal with nodes attribute. In this paper we propose a new method whic...

Graph clustering techniques are very useful for detecting densely connected groups in large graphs. Many existing graph clustering methods mainly focus on the topological structure, but ignore the vertex properties. Existing graph clustering methods have been recently extended to deal with nodes attribute. First we motivate the interest in the stud...

Attributed network models have seen an increasing success in recent years, thanks to their informative power and to their ability to model complex networked relations that characterize most real-world phenomena. Their use has been attractive to communities in different disciplines such as computer science, physics, social science, as well as in int...

Collaborative filtering is a well-known technique for recommender systems. Collaborative filtering models use the available preferences of a group of users to make recommendations or predictions of the unknown preferences for other users. Collaborative filtering suffers from the data sparsity problem when users only rate a small set of items which...

In this paper we propose a graph-coarsening approach that aims to speed-up the execution time of graph-based tag recommenders in large-scale folksonomies. A community detection algorithm in multiplex networks is applied for coarsening the hypergraph depicting a folksonomy. Experiments on real datasets show the validity of the approach.

In this paper we propose a new graph-based tag recommendation approach. The approach is structured into an offline step and an online one. Offline, the hypergraph depicting the history of tags assignment by users to resources is abstracted. On online, for a given target user and a resource, we first compute the set of recommended abstract tags (i.e...

"Acte du 1e atelier de Fouille de Grands Graphes", AFGG@EGC13, edit 13e Journées Francophones Extraction et Gestion des Connaissances, EGC’13, 60 pages, 52 pages, 29 janvier - 02 février 2013, Toulouse, France.

"Acte du 2e atelier de Fouille de Grands Graphes", AFGG@EGC14, edit 14e Journées Francophones Extraction et Gestion des Connaissances, EGC’14, 60 pages, 28 au 31 janvier 2014, Rennes, France.

This chapter presents the problem of link prediction in complex networks. It provides general description, formal definition of the problem and applications. It gives a state-of-art of various existing link prediction approaches concentrating more on topological approaches. It presents the main challenges of link prediction task in real networks. T...

Recommendation systems provide the facility to understand a person’s taste and find new, desirable content for them based on aggregation between their likes and rating of different items. In this paper, we propose a recommendation system that predict the note given by a user to an item. This recommendation system is mainly based on unsupervised top...

Multiplex network model has been recently proposed as a mean to capture high level complexity in real-world interaction networks. This model, in spite of its simplicity, allows handling multi-relationnal, heterogeneous, dynamic and even attributed networks. However, it requiers redefining and adapting almost all basic metrics and algorithms general...

Ensemble clustering approaches have been recently applied, in a variety of ways, in order to enhance the quality and/or the execution time of community detection tasks. The quality gain that can be obtained from applying ensemble approaches is known to be tightly linked to both quality and diversity of the applied clusterings. However, most of exis...

Performance of cluster ensemble approaches is now known to be tightly related to both quality and diversity of input base clusterings. Cluster ensemble selection (CES) refers to the process of filtering the raw set of base clusterings in order to select a subset of high quality and diverse clusterings. Most of existing CES approaches apply one inde...

In this work we present a new approach for co-authorship link prediction based on leveraging information contained in general bibliographical multiplex networks. A multiplex network is a graph defined over a set of nodes linked by different types of relations. For instance, the multiplex network we are studying here is defined as follows : nodes re...

Multiplex network is an emergent model that has been lately proposed in order to cope with the complexity of real-world networks. A multiplex network is defined as a multi-layer interconnected graph. Each layer contains the same set of nodes but interconnected by different types of links. This rich representation model requires to redefine most of...

Unfolding the community structure of complex networks is still to be one of the most important tasks in the field of complex network analysis. However, in many real settings, we seek to uncover the community of a given node rather than partitioning the whole graph into communities. A main trend in the area of ego-centred community identification co...

In this paper we propose a new approach for efficiently identifying local communities in complex networks. Most existing approaches are based on applying a greedy optimization process guided by a given objective function. Different objective functions have been proposed in the scientific literature, each capturing some specific feature of desired c...

This chapter develops a new type of Knowledge Organization System (KOS) based on the tagging of digital documents in organizations. The system, called hypertagging, is mainly based on the principle of faceted classification. The chapter presents an overview of the latest developments in the field of tagging recommender systems. The propagation of a...

In this work, we present an original seed-centric algorithm for community detection. Instead of expanding communities around selected seeds as most of existing seed-centric approaches do, we propose applying an ensemble clustering approach to different network partitions derived from local communities computed for each seed. Local communities are t...

Leader-driven community detection algorithms (LdCD hereafter) constitute a new trend in devising algorithms for community detection in large-scale complex networks. The basic idea is to identify some particular nodes in the target network, called leader nodes, around which local communities can be computed. Being based on local computations, they a...

In this paper we present an original approach for community detection in complex networks. The approach belongs to the family of seed-centric algorithms. However, instead of expanding communities around selected seeds as most of existing approaches do, we explore here applying an ensemble clustering approach to different network partitions derived...

Seed-centric algorithms constitue an emerging trend in the hot area of community detection in complex networks. The basic idea underlaying these approaches consists on identifying special nodes in the target network, called seeds, around which communities can then be identified. Different algorithms adopt different seed definitions and apply differ...

In this paper we present an original approach for community detection in complex networks. The approach belongs to the family of seed-centric algorithms. However, instead of expanding communities around selected seeds as most of existing approaches do, we explore here applying an ensemble clustering approach to different network partitions derived...

In this paper we propose a new topological approach for link prediction in dynamic complex networks. The proposed approach applies a supervised rank aggregation method. This functions as follows: first we rank the list of unlinked nodes in a network at instant t according to different topological measures (nodes characteristics aggregation, nodes n...

In this work, we explore applying a link prediction approach to tag recommendation in broad folksonomies. The original idea of the approach is to mine the dynamic of the tagging activity in order to compute the most suitable tag for a given user and a given resource. The tagging history of each user is modeled by a temporal sequence of bipartite gr...

In this work we propose a new efficient algorithm for communities construction based on the idea that a community is animated by a set of {\em leaders} that are followed by a set of nodes. A node can follow different leaders animating different communities. The algorithm is structured into two mains steps: identifying nodes in the network playing t...

One of the primary goals of tag recommendation approaches is to deal with the problem of ambiguity of tags in a folksonomy by helping users to select the most appropriate tag to annotate a resource. We propose in this work, an original approach for tag recommendation applying a link prediction using supervised machine learning. Given a user (target...

In this paper we present Casep2: a hybrid neuro-symbolic system combining case-based reasoning (CBR) and artificial neural networks that aims at clustering and classifying users' behavior in an e-commerce site. A user behavior is represented by a sequence of visited web pages, in a session. Each registered behavior is associated to one of the follo...

This work copes with the problem of link prediction in large-scale two-mode social networks. Two variations of the link prediction tasks are studied: predicting links in a bipartite graph and predicting links in a unimodal graph obtained by the projection of a bipartite graph over one of its node sets. For both tasks, we show in an empirical way, t...

This paper presents a new approach in the management of mobile ad hoc networks. Our alternative, based on mobile agent technology, allows the design of mobile centralized server in ad hoc network, where it is not obvious to think about a centralized management, due to the absence of any administration or fixed infrastructure in these networks. The...

In this work we tackle the problem of link prediction in co-authoring network. We apply a topological dyadic supervised machine learning approach for that purpose. A co-authoring network is actually obtained by the projection of a two-mode graph (an authoring graph linking authors to publications they have signed) over the authors set. We show that...

In this paper, we present COBRAS: a CBR-based peer-to-peer bibliographical reference recommender system. The system allows a group of like-minded people to share their bibliographical data in an implicit and intelligent way. The system associates a software agent with each user. Agents are attributed three main skills: a) detecting the associated u...

This paper presents a new approach in the management of mobile ad hoc networks. Our alternative, based on mobile agent technology, allows the designing of mobile centralized servers in ad hoc network, where it is not obvious to think about a centralized management, due to the absence of any administration or fixed infrastructure in these networks....

Edition d'actes de l'atelier AGS: Apprentissage et Graphes pour les Systèmes complexes (60 pages)

Mobility and the new mode of communication used in ad hoc mobile networks cause problems proper to a mobile environment. In this paper, we study the localization problem of mobile agents in ad hoc mobile networks; an object which has a double mobility is to be located, first, in relation to the mobility of its physical support in a dynamic environm...

This paper provides a state of art survey of works using social network analysis (SNA) techniques for improving performances of an on-line help desk. The processing of a client request Qc can be achieved following oe ofg the following scenarios: 1) sending Qc to the help desk. 2) sending QC to other clients that may have an appropriate answer., and...

In this paper we describe a peer-to-peer approach that ails at allowing a group of like-minded people to share relevant documents in an implicit way. We suppose that user save their documents in a local user-defined hierarchy. the association between documents and hierarchy nodes (or folders) is used by a supervised hybrid neural-CBR classifier in...

Nous présentons dans ce chapitre un système baptisé CASEP2, qui combine le raisonnement à partir de cas (RàPC) avec des réseaux de neurones (RN), pour le classement (ou la prédiction) à partir de séquences. Dans CASEP2, un cas est modélisé par une matrice de covariance dynamique qui prend en compte la distribution des états de la séquence dans l'es...

In this paper, we describe a cooperative P2P bibliographical data management and recommendation system (COBRAS). In COBRAS, each user is assisted by a personal software agent that helps her/him to manage bibliographical data and to recommend new bibliographical references that are known by peer agents. Key problems are: – how to obtain relevant ref...