Samira Sadaoui

Winner Determination in Combinatorial Reverse Auctions

Since commercially efficient, combinatorial auctions are getting more interest than traditional auctions. However, winner determination problem is still one of the main challenges of combinatorial auctions. In this paper, we propose a new method based on genetic algorithms to address two important issues in the context of combinatorial reverse auctions: determining the winner(s) in a reasonable processing time and reducing the procurement cost. Indeed, not much work has been done using genetic algorithms to determine the winner(s) specifically for combinatorial reverse auctions. To evaluate the performance of our method, we conducted several experiments comparing our proposed method with another method related to determining winner(s) in combinatorial reverse auctions. The experiment results clearly demonstrate the superiority of our method in terms of processing time and procurement cost.

A trust-based service suggestion system using human plausible reasoning

Nowadays, there is a growing need to manage trust in open systems as they may contain untrustworthy service providers. Agent Trust Management (ATM) tries to address the problem of finding a set of the most trusted agents in multi agent systems. This paper presents ScubAA, a novel generic ATM framework based on the theory of Human Plausible Reasoning (HPR). For each user’s request, ScubAA determines a ranked list of the most trusted service agents, within the context of the request, and forwards the request to those trusted services only. ScubAA determines an agent’s degree of trust in terms of a single personalized value derived from several types of evidences such as user’s feedback, history of user’s interactions, context of the submitted request, references from third party users as well as from third party service agents, and structure of the society of agents. ScubAA is able to utilize more trust evidences towards a more accurate value of trust. We also propose a function to figure out how similar two users are in a given context. We apply the proposed HPR-based ATM framework to the domain of Web search. The resulting ATM system recommends to the user a list of the most trusted search engines ranked according to the retrieval precision of documents returned in response to the user’s query as well as the degree of trust of the search engines have gained by interacting with other related users within the context of the query. In addition, we conduct a statistical analysis of ScubAA based on ANOVA and by using a data set of forty queries in different domains. This analysis clearly reveals that ScubAA is able to successfully assess the trustworthiness of service agents.

An Empirical Analysis of Imbalanced Data Classification

SVM has been given top consideration for addressing the challenging problem of data imbalance learning. Here,we conduct an empirical classification analysis of new UCI datasets that have dierent imbalance ratios, sizes andcomplexities. The experimentation consists of comparing the classification results of SVM with two other popularclassifiers, Naive Bayes and decision tree C4.5, to explore their pros and cons. To make the comparative exper-iments more comprehensive and have a better idea about the learning performance of each classifier, we employin total four performance metrics: Sensitive, Specificity, G-means and time-based eciency. For each benchmarkdataset, we perform an empirical search of the learning model through numerous training of the three classifiersunder dierent parameter settings and performance measurements. This paper exposes the most significant resultsi.e. the highest performance achieved by each classifier for each dataset. In summary, SVM outperforms the othertwo classifiers in terms of Sensitive (or Specificity) for all the datasets, and is more accurate in terms of G-meanswhen classifying large datasets.

A dynamic stage-based fraud monitoring framework of multiple live auctions

Monitoring the progress of auctions for fraudulent bidding activities is crucial for detecting and stopping fraud during runtime to prevent fraudsters from succeeding. To this end, we introduce a stage-based framework to monitor multiple live auctions for In-Auction Fraud (IAF). Creating a stage fraud monitoring system is different than what has been previously proposed in the very limited studies on runtime IAF detection. More precisely, we launch the IAF monitoring operation at several time points in each running auction depending on its duration. At each auction time point, our framework first detects IAF by evaluating each bidder’s stage activities based on the most reliable set of IAF patterns, and then takes appropriate actions to react to dishonest bidders. We develop the proposed framework with a dynamic agent architecture where multiple monitoring agents can be created and deleted with respect to the status of their corresponding auctions (initialized, completed or cancelled). The adoption of dynamic software architecture represents an excellent solution to the scalability and time efficiency issues of IAF monitoring systems since hundreds of live auctions are held simultaneously in commercial auction houses. Every time an auction is completed or terminated, the participants’ fraud scores are updated dynamically. Our approach enables us to observe each bidder in each live auction and manage his fraud score as well. We validate the IAF monitoring service through commercial auction data. We conduct three experiments to detect and react to shill-bidding fraud by employing datasets acquired from auctions of two valuable items, Palm PDA and XBOX. We observe each auction at three-time points, verifying the shill patterns that most likely happen in the corresponding stage for each one.

Data Mining Techniques and Preference Learning in Recommender Systems

The importance of implementing recommender systems has significantly increased during the last decade. The majority of available recommender systems do not offer clients the ability to make selections based on their choices or desires. This has motivated the development of a web based recommender system in order to recommend products to users and customers. The new system is an extension of an online application previously developed for online shopping under constraints and preferences. In this work, the system is enhanced by introducing a learning component to learn user preferences and suggests products based on them. More precisely, the new component learns from other customers’ preferences and makes a set of recommendations using data mining techiques including classification, association rules and cluster analysis techniques. The results of experimental tests, conducted to evaluate the performance of this component when compiling a list of recommendations, are very promising.

Generalization for component reuse

Specification reuse is more promising than code reuse since the formal semantics makes it possible for tools to understand the reusable components and ensures their correctness. One method for enhancing the reusability of existing components is generalization which creates generic components by parameterizing specific ones. Combining formal specifications and reusable components is a promising way to solve the software crisis.Component generalization is the abstraction of an auxiliary part of a specification into a more general parameter. Therefore, a major difficulty during the generalization is determining the appropriate level of abstraction (or generality). In this paper, through a simple example, we present the syntactic and semantic generalization algorithms based on algebraic specifications, and illustrate how to control the level of abstraction in generic components using the categorized constructors.

An approach to solve winner determination in combinatorial reverse auctions using genetic algorithms

Nowadays, winner determination problem is one of the main challenges in the domain of real-time applications such as combinatorial reverse auctions. To determine the winner(s) in combinatorial reverse auctions, in our previous work, we have proposed a Genetic Algorithm (GA)-based method and have demonstrated its superiority in terms of processing time and procurement cost. One of the main drawbacks of traditional GA-based solutions is their inconsistency in different runs. In this paper, we perform a statistical-based experiment that reveals that our proposed method is not affected by the inconsistency issue. In addition, we show two other features of our GA-based method: (1) the quality of the solution improves over generations, and (2) the any-time behavior.

Winner Determination in Multi-attribute Combinatorial Reverse Auctions

Winner(s) determination in online reverse auctions is a very appealing e-commerce application. This is a combinatorial optimization problem where the goal is to find an optimal solution meeting a set of requirements and minimizing a given procurement cost. This problem is hard to tackle especially when multiple attributes of instances of items are considered together with additional constraints, such as seller’s stocks and discount rate. The challenge here is to determine the optimal solution in a reasonable computation time. Solving this problem with a systematic method will guarantee the optimality of the returned solution but comes with an exponential time cost. On the other hand, approximation techniques such as evolutionary algorithms are faster but trade the quality of the solution returned for the running time. In this paper, we conduct a comparative study of several exact and evolutionary techniques that have been proposed to solve various instances of the combinatorial reverse auction problem. In particular, we show that a recent method based on genetic algorithms outperforms some other methods in terms of time efficiency while returning a near to optimal solution in most of the cases.

Constraint and Qualitative Preference Specification in Multi-Attribute Reverse Auctions

In the context of Multi-Attribute and Reverse Auctions MARAs, two significant problems need to be addressed: 1 specifying precisely the buyers requirements about the attributes of the auctioned product, and 2 determining the winner accordingly. Buyers are more comfortable in expressing their preferences qualitatively, and there should be an option to allow them describes their constraints. Both constraints and preferences may be non-conditional and conditional. However for the sake of efficiency, it is more suitable for MARAs to process quantitative requirements. Hence, there is a remaining challenge to provide the buyers with more facilities and comfort, and at the same time to keep the auctions efficient. To meet this challenge, we develop a MARA system based on MAUT. The proposed system takes advantage of the efficiency of MAUT by transforming the qualitative requirements into quantitative ones. Another benefit of our system is the complete automation of the bid evaluation since it is a really difficult task for buyers to determine quantitatively all the weights and utility functions of attributes, especially when there is a large number of attributes. The weights and utility functions are produced based on the qualitative preferences. Our MARA looks for the outcome that satisfies all the constraints and best satisfies the preferences. We demonstrate the feasibility of our system through a 10-attribute reverse auction involving many constraints and qualitative preferences.

Agent Trust Management Based on Human Plausible Reasoning: Application to Web Search

In open systems, different service providers can join and leave at any time. Multi Agent Systems (MASs) are being used more and more as the basis of open systems. Although openness brings a huge opportunity for different systems to operate in a decoupled and autonomous manner, it can introduce untrustworthy agents into the society. For this purpose, Agent Trust Management (ATM) methods have been proposed to try to eliminate this defect. This paper presents a general framework for managing trust in MASs based on the theory of Human Plausible Reasoning (HPR). The goal of the proposed framework is to determine for each user a ranked list of trusted agents and to find newer possible trust relationships between users and agents. We use the HPR certainty parameters to define how trustworthy each agent is in the list. We measure the agent trust according to two metrics: the direct interaction rating and third-party references. For each user, a third party is any other user with whom a HPR relationship exists. We aggregate the direct interaction rating value and the reputation values of third parties to achieve a single quantitative value for the trust. This value is then used to rank the agents. We apply our HPR-based ATM framework to the domain of Web search. The resulting ATM system provides the user a list of trusted search engines ranked according to the reputation the search engine has gained by interacting with other related users as well as the retrieval precision of pages returned in response to the users query.