Aziz Barbar

A New Heuristic Algorithm for the 3D Bin Packing Problem

3D bin packing is a classical NP-hard (Nondeterministic Polynomial-time hard) problem where a set N of 3D boxes is to be packed in a minimum number of containers (bins). 3D bin packing is used in many industrial applications; hence computer scientists are challenged in designing practical and efficient approaches for the problem. This paper presents a new heuristic algorithm called Peak Filling Slice Push (PFSP) for 3D bin packing. The algorithm recursively divides the container into smaller slices and then fills each slice with boxes before pushing them to minimize the wasted space. Experimental results showed significant performance improvements over other current approaches.

Support vector machines: A distance-based approach to multi-class classification

One of the main tasks sought after with machine learning is classification. Support vector machines are one of the widely used machine learning algorithms for data classification. SVMs are by default binary classifiers, extending them to multi-class classifiers is a challenging on-going research problem. In this paper, we propose a new approach to constructing the multi-class classification function, where the structure and properties of the support vectors are exploited without altering the training procedure. Our contribution is based on the insight that one is not restricted to using the hyperplane-based decision function, resulting from the mathematical optimization problem. The proposed classification procedure considers the notion of distance between vectors in feature space. We show how, under the assumption of a normalized kernel, the distance between two vectors in feature space can be expressed solely in terms of their inner product. We apply both the original and proposed methods on synthetic datasets in a simulation setting, and then we argue that the proposed distance-based method represents a more rigorous and intuitive measure than the traditional hyperplane-based method.

Cube Objects with Functionality Facets Enhancer Entities - COFFEE

The interaction between end-users and computers is attained through the operating systempsilas user interface. 2D graphical user interfaces along with the WIMP metaphor have long been governing this interaction. A new, 3D approach is introduced in this paper, whereby new metaphors are established. First, related works in the field are discussed. Then, COFFEE (Cube Objects with Functionality Facets Enhancer Entities) is introduced with the new metaphors and 3D graphical enhancements. Then, the paper is concluded.

A distance-based approach for database re-engineering

We present the UQoRE method which supports database reverse engineering by using a data mining technique. Our approach proposes to mine user queries collected on the database in order to extract a specific distance-based similarity measure between two attributes. Unlike previous propositions which were based on the observation of query structures, we focus on the exploration of a query repository. The objective is to extract semantic links which do not appear obviously from other available information sources and are suggested implicitly in queries by expert users who have acquired a good knowledge of the application context. We focus mainly on the problem of attribute similarity that is quite important for the following reengineering steps according to context-based measures without taking into consideration the naming policy used by database designers.

BlueKey - A Bluetooth Secure Solution for Accessing Personal Computers

A major realm of security breach for today’s users is unprivileged access, modification or sometimes forgery of critical business, or user information. Existing computer locking/unoking methods serve as an intermediate barrier against unethical deeds. The proposed solution BlueKey is a software-based solution installed on Personal Computers (PC) that safely unlocks the PC by securing the Bluetooth communication channel between the user’s mobile device and his/her PC. BlueKey helps end-users in not typing their passwords every time they need to access their PCs. At the same level, this solution includes a mobile application that allows the owner to fully control his/her PC via Bluetooth, and runs a breach detector module with safety measures to protect the PC when it is locked. At the technical level, BlueKey is a platform free application written using Java programming language, fulfilling the Write Once Run Anywhere (WORA) concept. This system is built using Java Development Kit (JDK) and compiled using Java Virtual Machine (JVM), and runs with Java Runtime Environment (JRE). Alongside, the mobile application is developed using Java 2 Micro Edition (J2ME), which is compatible with Android, Symbian, and BlackBerry operating systems.

Stochastically Balancing Trees for File and Database Systems

With the constant improvement in data storage technologies, a new generation of indexing mechanisms is to be created to exploit the improvements in disk access speeds that were previously impractical. The self-balancing tree B-Tree, has long been the indexing structure of choice for reducing the amount of disk access at the expense of size of data block to be read or written. A new technique based on a dynamically growing multilevel list structure, which is stochastically balanced rather than self balanced, is discussed and compared to the B-Tree. An analogy between the technique and the structures is established to better compare the computational complexities.

P2P Queries Routing Using Distributed Knowledge

The growing interest in Peer-to-Peer systems (such as Gnutella) has inspired numerous research activities. The problem in a schema-based Peer-to-Peer (P2P) system is how to locate Peers relevant to a given query. Different methods proposed routing strategies of queries taking into account the P2P network at hand. In this paper, we propose an architecture, based on (Super-) Peers, and we focus on query routing. Our approach considers that (Super-) Peers having similar interests are grouped together for an efficient query routing method. In such groups, called Super-Super-Peers (SSP), Super-Peers submit queries that are often processed by members of this group. A SSP is a specific Super-Peer that contains knowledge about: 1. its Super-Peers and 2. The other SSP. Knowledge is extracted by using data mining techniques (e.g. decision tree algorithms) starting from queries of Peers that transit on the network. The advantage of this distributed knowledge is that, it avoids making semantic mapping, between heterogeneous data sources owned by (Super-)Peers, each time the system decides to route query to other (Super-)Peers. The set of SSP improves the robustness in queries routing mechanism and scalability in P2P Network. Compared with a baseline approach, our proposed architecture shows that data mining technique increase performance with respect to response time and precision.

Using data mining to minimize database reverse engineering constraints

In this paper we propose to use data mining techniques for database reverse engineering process. A crucial problem in this process concerns the discovery of similarity between attributes before constructing the conceptual model. The essence of our approach is to mine user queries collected on the database in order to extract specific similarity measure that we call distance between 2 attributes. Indeed most database reverse engineering methods are based on the observation of several sources which generally are the existing database schema, the data themselves and application programs including queries. Unlike previous propositions which analyze only the structure of joins in queries, the main idea of this paper is to exploit the large volume of information stored in queries in order to extract some semantic properties on attributes. Thus we propose to apply a data mining algorithm on a query base collected on the data. The objective is to extract semantic links that do not appear obviously in the schema or in the data and are suggested implicitly by expert users in their queries. In this paper, we focus mainly on the problem of attribute similarity which is quite important in database reverse engineering. We describe a method by which similarities between attributes are discovered according to context measures without taking into consideration the naming policy used by database designers.