Pavlos Delias

AgentAllocator: An Agent-Based Multi-criteria Decision Support System for Task Allocation

The multi-agent systems in Artificial Intelligence very often include task allocation problems. These problems are arduous to be modelled; therefore it is also difficult to end up with an optimal allocation plan. AgentAllocator is an easy to use, platform independent application, which implements a multi-criteria method to support the decision of Task Allocation. The decision maker is able to model the problem (according to his policy) through its inputs dialogs and employ the final solution proposed by the system. In this paper, both the theoretical background of AgentAllocator and the DSS itself are presented.

What agents can do in workflow management systems

Workflow management systems are an emerging category of information systems, currently under dynamic evolution. On the other hand software agents are a distinct research area as well as an emerging paradigm for information systems design and development. This paper tries to examine the integration of these two fields; reveal the stimulation and the advantages of such a mixing. Popular standards of the workflow management field are used to derive a classification scheme, which is exploited to map existing approaches. As a significant number of existing approaches is reviewed, a plethora of integration patterns are identified and grouped according to the proposed classification scheme. The overall goal of the paper is to clear the vague picture of the consolidation of workflow management systems and software agents and to provide an unifying framework for this intersected area.

A Multi-criteria Protocol for Multi-agent Negotiations

Negotiation processes are often characterized by conflicts of interests of the negotiating parts. However, it is possible to mitigate these conflicts, if we support the negotiation process with a well-structured model. This area of interest has large occupied the scientists of Group Decision Support Systems (GDSS) and particularly those who focus on the Negotiation Support Systems (NSS). In this paper, we propose an experimental multi-criteria prototype negotiation protocol which allows agents to follow a process, in order to end up with an optimal decision. The proposed model is able to estimate agents’ preferences and suggest a convenient solution.

User satisfaction and e-learning systems: Towards a multi-criteria evaluation methodology

E-learning systems permeate the modern informative approaches of education. As their number is growing up, a new demand arises. That demand is about evaluating those systems in order not to separate the “good” ones from the “bad” ones but to comprehend the factors that lead to an effective and useful e-learning system. This paper proposes a multi-criteria model that uses linear programming to measure a satisfaction index and to compute criteria weights. To evaluate the e-learning system, we use a set of fifteen sub-criteria that corresponds to three main criteria. In addition, there is a crest-question that counts the overall satisfaction. In this study, we focused on the multi-criteria methodology, but we also included our thoughts and ambitions on an expanding implementation through an e-learning system.

Discovering Process Horizontal Boundaries to Facilitate Process Comprehension

Automated discovery of a process model is a major task of Process Mining that means to produce a process model from an event log, without any a-priori information. However, when an event log contains a large number of distinct activities, process discovery can be real challenging. The goal of this article is to facilitate process discovery in such cases when a process is expected to contain a large set of unique activities. To this end, this article proposes a clustering approach that recommends horizontal boundaries for the process. The proposed approach ultimately partitions the event log in a way that human interpretation efforts are decomposed. In addition, it makes automated discovery more efficient as well as effective by simultaneously considering two quality criteria: informativeness and robustness of the derived groups of activities. The authors conducted several experiments to test the behavior of the algorithm under different settings, and to compare it against other techniques. Finally, they provide a set of recommendations that may help process analysts during the process discovery endeavor.

Optimizing differentiation and commonality levels among models in car line-ups

We implement a swarm intelligence mechanism to design optimal car lines.We aim to optimize differentiation and commonality levels among models in the line.Our mechanism utilizes stated preference data derived from a conjoint experiment.A prototype system is also developed to facilitate managerial decision making.Insights are provided into how new and existing car models must be combined. The product life cycle of cars is becoming shorter and carmakers constantly introduce new or revised models in their lines, tailored to their customer needs. At the same time, new car model design decisions may have a substantial effect on the cost and revenue drivers. For example, although a new car model configuration with component commonality may lower manufacturing cost, it also hinders increased revenues that could have been achieved through product differentiation. This paper develops and illustrates a state of the art, nature-inspired approach, to design car lines that optimize the degree of differentiation vs commonality among models in the line. More specifically, we apply a swarm intelligence mechanism to stated preference data derived from a large-scale conjoint experiment that measures consumer preferences for passenger cars in a sample of 1164 individuals. The proposed two-step methodology is also incorporated into a prototype system, which has been developed in an attempt to facilitate managerial decision making. Our approach provides interesting insights into how new and existing car models can be combined in a product line and identifies the desired balance between differentiation and commonality levels among models within a product line, which elevates customer satisfaction.

A genetic approach for strategic resource allocation planning

Enterprises often implement a measurement system to monitor their march towards their strategic goals. Although this way it is possible to assess the progress of each goal, there is no structured way to reconsider resource allocation to those goals and to plan an optimal (or near optimal) allocation scheme. In this study we propose a genetic approach to match each goal with an autonomous entity (agent) with a specific resource sharing behavior. The overall performance is evaluated through a set of functions and genetic algorithms are used to eventuate in approximate optimal behavior’s schemes. To outline the strategic goals of the enterprise we used the balanced scorecard method. Letting agents deploy their sharing behavior over simulation time, we measure the scorecard’s performance and detect distinguished behaviors, namely recommendations for resource allocation.

Exploring the factors affecting Greek people to make vacation in Greece: A preliminary analysis

Abstract This paper reports results of a study aiming at exploring the factors affecting Greek people to make vacation in Greece. Through the study i) the main factors influencing Greek people in their choice about the tourist destination ii) consumers with similar buying behaviour were classified into groups and iii) each group of consumers was profiled according to their demographic characteristics and their preferences regarding their holidays. Field interviews were conducted in a randomly selected sample consisting of 102 people in Greece in November 2013. Principal Components Analysis (PCA) was performed to identify the main factors that affect Greek people to make vacation in Greece. These factors were: (a) vacation activities (b) natural environment and entertainment (c) travel connections and (c) culture. Hierarchical and non hierarchical cluster techniques were employed to classify consumers with similar consumption behaviour. Three groups of tourists were identified: (a) those interested in the vacation activities, (b) those interested in cultural issues, (c) those interested in travel connection. Discriminant Analysis was performed to assess how the identified main factors affecting Greek people to have holidays in Greece through PCA, could predict cluster membership. A non parametric test was used to profile each group of tourists according to their demographic characteristics.

Profiling Tourists who have Holidays in the Region of Eastern Macedonia and Thrace in Greece

Abstract This paper aims to profile the tourists who have holidays in the Region of Eastern Macedonia and Thrace (REMTh) in Greece regarding the factors affecting them in choosing their travel destination and sources of information and demographic characteristics. In particular the main factors influencing tourists to choose REMTh as tourist destination were identified. Furthermore tourists with similar buying behaviour were classified into groups; each group of tourists was profiled according to their preferences regarding their holidays and their demographic characteristics. Field interviews were conducted in a randomly selected sample consisting of 265 people in REMTh in summer 2013. Principal Components Analysis (PCA) was performed to identify the main factors that affect tourists to make vacation in REMTh. These factors were: (a) Travel connections and tourist infrastructures (b) natural environment (c) vacation activities, (d) entertainment, (e) culture and (f) value for money. Hierarchical and non hierarchical cluster techniques were employed to classify tourists with similar behaviour. Three groups of tourists were identified: (a) those interested in the vacation activities, entertainment and culture (b) opportunists, (c) those interested in natural environment and (d) those interested in travel connections, tourist infrastructures and value for money. Discriminant Analysis was performed to assess how the identified main factors affecting tourists to have holidays in REMTh through PCA, could predict cluster membership. A non parametric test was used to profile each group of tourists regarding their holidays and their demographic characteristics.

Manual Intervention and Statefulness in Agent-Involved Workflow Management Systems

Lack of adaptability within WorkFlow Management Systems (WFMS) has been early identified as one of their limitations. WFMS suffer from disadvantages such as not supporting the dynamic incorporation/modification of process models and poor adaptability of process models at runtime. The static workflow definition and its passive interpretation does not allow WFMS to demonstrate flexible behavior and to deal with real-life situations, such as fast changing customer requirements and enterprise goal shifts. In this work we propose the design and development of two features (manual intervention and statefulness), which are expected to tackle this limitation. Our work considers and agent-based environment for the WFMS implementation.