Aristotle University Of Thessaloniki (AUTH)
Εξόρυξη δεδομένων για τη δυναμική ενσωμάτωση γνώσης σε πολυπρακτορικά συστήματα
Abstract
dc:descriptionThe task-oriented nature of Data Mining has already been dealt successfully with the employment of Intelligent Agent systems that distribute tasks, collaborate and synchronize in order to reach their ultimate goal, the extraction of knowledge. A number of sophisticated multi-agent systems that perform data mining have been developed, proving that Agent Technology can indeed be used in order to solve data mining problems. Looking into the opposite direction though, knowledge extracted through data mining has not yet been exploited on multi-agent systems. The inductive nature of data mining imposes logic limitations and hinders the application of the extracted knowledge on such kind of deductive systems. If these limitations are overcome, the synergy of the technologies may become feasible. Within the context of this thesis, a unified methodology for building agents and multi-agent systems is presented. Agents and agent systems developed following the proposed methodology have the ability of dynamically incorporating knowledge, extracted by the use of data mining techniques on application and agent data. Three types of knowledge are identified: a) knowledge extracted by the use of data mining techniques on historical application data, b) knowledge extracted by the use of data mining techniques on agent behavior data and, c) knowledge extracted by the use of evolutionary data mining techniques on multi-agent communities. The four dominant data mining techniques are studied, in order to investigate their applicability potential: Clustering for grouping, Classification for categorization and prediction, Association Rule Extraction for correlation discovery and Genetic algorithms for self-organization. In order to prove the validity of the proposed approach, Data Miner, a tool for embedding knowledge extracted by the use of data mining techniques has been developed and is presented (Chapter 4). For each knowledge type identified, a representative test case has been developed and is described in detail (Chapters 5–7). The presented thesis also deals with retraining, the improvement of the reasoning models of agents, by the application of data mining on new, up-to-date data (Chapter 8). Finally, conclusions drawn and future work are summarized in Chapter 9.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2004
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Συμεωνίδης, Ανδρέας
- Symeonidis, Andreas
Subjects
dc:subject × 18- Εξόρυξη δεδομένων
- Πολυπρακτορικά συστήματα
- Δυναμική εκπαίδευση πρακτόρων
- Γνωσιακά μοντέλα
- Υποστήριξη αποφάσεων
- Συμπεριφορά πρακτόρων
- Κοινωνίες πρακτόρων
- Data mining
- Multi-agent systems
- Dynamic agent training
- Knowledge model
- Decision support
- Agent behaviour
- Agent communities
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Engineering and Technology
- Electrical Engineering, Electronic Engineering, Information Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/15173