Imen Boukhris

Evidential Item-Based Collaborative Filtering

Recommender Systems (RSs) in particular the collaborative filtering approaches have reached a high level of popularity. These approaches are designed for predicting the user’s future interests towards unrated items. However, the provided predictions should be taken with restraint because of the uncertainty pervading the real-world problems. Indeed, to not give consideration to such uncertainty may lead to unrepresentative results which can deeply affect the predictions’ accuracy as well as the user’s confidence towards the RS. In order to tackle this issue, we propose in this paper a new evidential item-based collaborative filtering approach. In our approach, we involve the belief function theory tools as well as the Evidential K-Nearest Neighbors (EKNN) classifier to deal with the uncertain aspect of items’ recommendation ignored by the classical methods. The performance of our new recommendation approach is proved through a comparative evaluation with several traditional collaborative filtering recommenders.

Proactive and reactive e-government services recommendation

AbstractGovernmental portals designed to provide electronic services are generally overloaded withn information that may hinder the effectiveness of e-government services. This paper proposes a new framework to supply citizens with adapted content and personalized services that satisfy their requirements and fit with their profiles in order to guarantee nuniversal access to governmental services. The proposed reactive and proactive solutions combine several nrecommendation techniques that use different data sources i.e., citizen profile, social media databases, citizen’s feedback databases and service databases. It is shown that recommender systems provide citizens with accessible personalized e-government services.

Dealing with external actions in belief causal networks

Abstract Graphical models are efficient and simple ways to represent dependencies between variables. We introduce in this paper the so-called belief causal networks where dependencies are uncertain causal links and where the uncertainty is represented by belief masses. Through these networks, we propose to represent the results of passively observing the spontaneous behavior of the system and also evaluate the effects of external actions. Interventions are very useful for representing causal relations, we propose to compute their effects using a generalization of the “do” operator. Even if the belief chain rule is different from the Bayesian chain rule, we show that the joint distributions of the altered structures to graphically describe interventions are equivalent. This paper also addresses new issues that are arisen when handling interventions: we argue that in real world applications, external manipulations may be imprecise and show that they have a natural encoding under the belief function framework.

Similarity Analysis of Protein Binding Sites: A Generalization of the Maximum Common Subgraph Measure Based on Quasi-Clique Detection

Protein binding sites are often represented by means of graphs capturing their most important geometrical and physicochemical properties. Searching for structural similarities and identifying functional relationships between them can thus be reduced to matching their corresponding graph descriptors. In this paper, we propose a method for the structural analysis of protein binding sites that makes use of such matching techniques to assess the similarity between proteins independently of sequence or fold homology. More specifically, we propose a similarity measure that generalizes the commonly used maximum common subgraph measure in two ways. First, using algorithms for so-called quasi-clique detection, our measure is based on maximum ‘approximately’ common subgraphs, a relaxation of maximum common subgraphs which is tolerant toward edge mismatches. Second, instead of focusing on equivalence, our measure is a compromise between a generalized equivalence and an inclusion measure. An experimental study is presented to illustrate the effectiveness of the method and to show that both types of relaxation are useful in the context of protein structure analysis.

Crowd Label Aggregation Under a Belief Function Framework

Crowdsourcing emerged as an efficient human-powered concept to tackle the problem of labeling complex tasks that computer programs still cannot solve. Amazon’s Mechanical Turk is one of the most popular platforms that allows to gather labels from human workers. These labels are then aggregated in order to estimate the true labels. Considering that not all labelers are experts, their answers may be imperfect and consequently unreliable. In this paper, we propose a novel label aggregation method based on the belief function theory. The proposed method grants a strong framework that does not only allow to reliably aggregate imperfect labels but also to integrate labelers expertise for more accurate results. To demonstrate the effectiveness of the proposed method, experiments are conducted on real datasets. The results show that our method is a promising solution in the crowd labeling domain.

Decision Model for Policy Makers in the Context of Citizens Engagement

Citizens’ engagement is considered as one of the important dimensions for the development of smart cities since, in the vision of a city of the future (smart city), citizens will be more and more involved in the decision-making process of different issues related to the development of a city. In this context, policy makers face a decision problem where they have to integrate a new dimension, which is the voice of the citizens’ decision. This article proposes a tool based on multicriteria decision making methods to provide decision makers with the best alternative(s) that are based on citizens’ opinions. In order to tackle the potential interdependencies between criteria and also between alternatives in the selection process, we apply a hybrid model integrating the analytical network process and an extended version of technique for order performance by similarity to ideal solution to support group decision-making. The proposed model is applied in the context of participatory budgeting (PB) where citizens decide on the projects in which the money can be invested. This process is complex since it encompasses multiple interdependent criteria that may be conflicting with each other and that are used to take decisions. To illustrate our approach, we will apply the proposed technique for the case study of La Marsa, a city in the north of the capital Tunis (Tunisia) that adopted, since 2014, a PB strategy in which citizens proposed alternatives on how an amount of money can be used to lighten specific streets in the city.

The Link Prediction Problem under a Belief Function Framework

Link prediction is a key research area in social network analysis that enables to understand how social networks evolve over time. It involves predicting the links that may appear in the future based on a snapshot of the social network. Various techniques addressing this problem exist but most of them deal with it under a certain framework. Yet, complete information about the social network of interest is frequently not available as knowledge about the nodes and edges may be partial and incomplete, hence any analysis approach must handle uncertainty in the prediction task. In this paper, we examine the link prediction problem in uncertain social networks by adopting the theory of belief functions. Firstly, a new graph-based model for social networks that encapsulates the uncertainties in the links structures is proposed. Secondly, we use the assets of the belief function theory for combining pieces of evidence induced from different sources and decision making to propose a novel approach for predicting future links through information fusion of the neighboring nodes. The performance of the new method is validated on a real world social network graph of Facebook friendships.

Evidential Link Prediction Based on Group Information

Link prediction has become a common way to infer new associations among actors in social networks. Most existing methods focus on the local and global information neglecting the implication of the actors in social groups. Further, the prediction process is characterized by a high complexity and uncertainty. In order to address these problems, we firstly introduce a new evidential weighted version of the social networks graph-based model that encapsulates the uncertainty at the edges level using the belief function framework. Secondly, we use this graph-based model to provide a novel approach for link prediction that takes into consideration both groups information and uncertainty in social networks. The performance of the method is experimented on a real world social network with group information and shows interesting results.

Community detection for graph-based similarity

We present a novel approach for similarity assessment between graphs.We formulate a new similarity measure that contributes to the graph matching problem.Our approach gives the best results in terms of accuracy and computation time. This paper addresses the problem of similarity assessment between node-labeled and edge-weighted graphs representing protein binding pockets. A novel approach is proposed for predicting the functional family of proteins on the basis of the properties of their binding pockets using graphs as models to depict their geometry and physicochemical composition without information loss. State of the art graph similarity measure based on the maximum common subgraph is relaxed by the use of an another concept: the so-called community, or in our context, the maximum densest common community MDCC, which is used as an almost common subgraph. The latter is more convenient since it allows to take into account the flexible nature of proteins on the 3D-level. With our approach, tolerance towards noise and structural variation is increased. Furthermore, the MDCC is detected with low computation time. The performance of our method is validated on real world data.

Representing belief function knowledge with graphical models

Belief function theory is an appropriate framework to model different forms of knowledge including probabilistic knowledge. One simple and efficient way to reason under uncertainty, is the use of compact graphical models, namely directed acyclic graphs. Therefore naturally, a question crosses the mind: If we deal with Bayesian belief knowledge does the network collapse into a Bayesian network? This paper attempts to answer this question by analyzing different forms of belief function networks defined with conditional beliefs defined either with a unique conditional distribution for all parents or a single conditional distribution for each single parent. We also propose a new method for the deconditionalization process to compute the joint distribution.