Habiba Drias

Cooperative bees swarm for solving the maximum weighted satisfiability problem

Solving a NP-Complete problem precisely is spiny: the combinative explosion is the ransom of this accurateness. It is the reason for which we have often resort to approached methods assuring the obtaining of a good solution in a reasonable time. In this paper we aim to introduce a new intelligent approach or meta-heuristic named “Bees Swarm Optimization”, BSO for short, which is inspired from the behaviour of real bees. An adaptation to the features of the MAX-W-SAT problem is done to contribute to its resolution. We provide an overview of the results of empirical tests performed on the hard Johnson benchmark. A comparative study with well known procedures for MAX-W-SAT is done and shows that BSO outperforms the other evolutionary algorithms especially AC-SAT, an ant colony algorithm for SAT.

An intrusion detection and alert correlation approach based on revising probabilistic classifiers using expert knowledge

Bayesian networks are important knowledge representation tools for handling uncertain pieces of information. The success of these models is strongly related to their capacity to represent and handle dependence relations. Some forms of Bayesian networks have been successfully applied in many classification tasks. In particular, naive Bayes classifiers have been used for intrusion detection and alerts correlation. This paper analyses the advantage of adding expert knowledge to probabilistic classifiers in the context of intrusion detection and alerts correlation. As examples of probabilistic classifiers, we will consider the well-known Naive Bayes, Tree Augmented Naïve Bayes (TAN), Hidden Naive Bayes (HNB) and decision tree classifiers. Our approach can be applied for any classifier where the outcome is a probability distribution over a set of classes (or decisions). In particular, we study how additional expert knowledge such as “it is expected that 80xa0% of traffic will be normal” can be integrated in classification tasks. Our aim is to revise probabilistic classifiers’ outputs in order to fit expert knowledge. Experimental results show that our approach improves existing results on different benchmarks from intrusion detection and alert correlation areas.

Multi-swarm bat algorithm for association rule mining using multiple cooperative strategies

Association Rule Mining (ARM) can be considered as a combinatorial problem with the purpose of extracting the correlations between items in sizeable datasets. The numerous polynomial exact algorithms already proposed for ARM are unadapted for large databases and especially for those existing on the web. Assuming that datasets are a large space search, intelligent algorithms was used to found high quality rules and solve ARM issue. This paper deals with a cooperative multi-swarm bat algorithm for association rule mining. It is based on the bat-inspired algorithm adapted to rule discovering problem (BAT-ARM). This latter suffers from absence of communication between bats in the population which lessen the exploration of search space. However, it has a powerful rule generation process which leads to perfect local search. Therefore, to maintain a good trade-off between diversification and intensification, in our proposed approach, we introduce cooperative strategies between the swarms that already proved their efficiency in multi-swarm optimization algorithm(Ring, Master-slave). Furthermore, we innovate a new topology called Hybrid that merges Ring strategy with Master-slave plan previously developed in our earlier work [23]. A series of experiments are carried out on nine well known datasets in ARM field and the performance of proposed approach are evaluated and compared with those of other recently published methods. The results show a clear superiority of our proposal against its similar approaches in terms of time and rule quality. The analysis also shows a competitive outcomes in terms of quality in-face-of multi-objective optimization methods.

Association Rule Mining Based on Bat Algorithm

In this paper, we propose a bat-based algorithm (BA) for association rule mining (ARM Bat). Our algorithm aims to maximize the fitness function to generate the best rules in the defined dataset starting from a specific minimum support and minimum confidence. The efficiency of our proposed algorithm is tested on several generic datasets with different number of transactions and items. The results are compared to FPgrowth algorithm results on the same datasets. ARM bat algorithm perform better than the FPgrowth algorithm in term of computation speed and memory usage,

A New Default Theories Compilation for MSP-Entailment

Handling exceptions represents one of the most important problems in Artificial Intelligence. Several approaches have been proposed for reasoning on default theories. This paper focuses on a possibilistic approach, and more precisely on the MSP-entailment (where MSP stands for Minimum Specificity Principle) from default theories which is equivalent to System P augmented by rational monotony. In order to make this entailment tractable from a computational point of view, we propose here a compilation of default theories with respect to a target compilation language. This allows us to provide polynomial algorithms to derive efficiently the MSP-conclusions of a compiled default theory. Moreover, the proposed compilation is qualified to be flexible since it efficiently takes advantage of any classical compiler and generally provides a low recompilation cost when updating a compiled default theory.

From Data Warehouse to Information Warehouse: Application to Social Media

Data and information are different constructs and Data become information when they supply a content from several elements, generating thus a meaning. From this observation, in this paper, we introduce the paradigm of information warehousing and explain our motivation. We propose a generic and original information warehouse architecture for the storage and analysis of all types of information sources such as scientific papers, press articles and social media. The described infrastructure is then illustrated for the case of Twitter where a multidimensional information model is defined. The collected information flow is analyzed using a data mining technique, which is the Apriori algorithm, to discover association rules that would reflect the topics discussed in the tweets collection. The obtained results are promising and confirm the potential of the proposed paradigm.

ULR-Discr: A New Unsupervised Approach for Discretization

In this work, we propose a novel unsupervized discretization method based on a Left to Right (LR) scanning technique, namely ULR-Discr. Its originality resides in the fact it uses fusion and division operations at the same time and among its strengths, we report two advantages. The first one consists in designing the algorithm by crossing the input stream in a single pass, and this way the time complexity is significantly reduced relatively to that of the previous works. The second is the possibility offered to provide easily any cut-point function to reach the desired effectiveness. To evaluate our method, extensive experiments were conducted on large datasets in order to undertake comparison with several classical discretization methods and recent ones.

Multi-objective Bat Algorithm for Mining Interesting Association Rules

Association rule mining problem attracts the attention of researchers inasmuch to its importance and applications in our world with the fast growth of the stored data. Association rule mining process is computationally very expensive because rules number grows exponentially as items number in the database increases. However, Association rule mining is more complex when we introduce the quality criteria and usefulness to the user. This paper deals with association rule mining issue in which we propose Multi-Objective Bat algorithm for association rules mining Known as MOB-ARM. With the aim of extract more useful and understandable rules. We introduce four quality measures of association rules: Support, Confidence, Comprehensibility, and Interestingness in two objective functions considered for maximization. A series of experiments are carried out on several well-known benchmarks in association rule mining field and the performance of our proposal are evaluated and compared with those of other recently published methods including mono-objective and multi-objective approaches. The outcomes show a clear superiority of our proposal in-face-of mono objective methods in terms generated rules number and rule quality. Also, The analysis also shows a competitive outcomes in terms of quality against multi-objective optimization methods.