Wolfgang Benn

Enabling integrative negotiations by adapitve software agents

One of the most important and complex aspects in Electronic Commerce is the automated negotiation between software agents. Most shopping assistance systems concentrate on price comparison agents or provide only the means to negotiating the price of articles. In our paper we present the model of an intelligent and adaptive software agent to support integrative negotiations of multiple interdependent properties of products. The agent is able to consider an arbitrary number of different properties whose values may be interdependent. Additionally, a ranking of the importance of the properties is maintained. Normalization of property values is an integral part of our model. The user does not need to transform desired values to abstract scales. Furthermore, normalization allows properties of different types to be the subject of one negotiation. Our agent can selectively act as a buyer or a seller of a product. The intelligent and adaptive behaviour throughout the negotiation is secured by a set of strategies, which determine the agents proposals and its reactions to its counterpart. The realization of these negotiation strategies is based on ideas adopted from the field of neural networks and vibrations theory.

Zugriff auf Datenbanken über das World Wide Web

Zusammenfassung  Datenbanksysteme gewinnen für das World Wide Web (WWW) zunehmend an Bedeutung, da sie die Organisation großer Informationsmengen sowie die transaktionsbasierte und sichere Verwaltung von Daten und Geschäftsvorgängen unterstützen. Um effizient auf Datenbanksysteme (DBS) in Internetanwendungen zugreifen zu können, müssen geeignete Übergänge zwischen WWW und DBS benutzt werden. Wir stellen anhand eines Forderungskatalogs die Eigenschaften aktueller Verfahren vor, ordnen sie den Architekturen der Informationsverarbeitung zu und diskutieren die Möglichkeiten weiterer Entwicklungen.Summary  The importance of Database Systems (DBS) for the World Wide Web (WWW) is increasing significantly, because they do not only support the organization of large amounts of information but also the transaction-based and secure management of data and business processes. In order to access database systems from Internet applications efficiently, appropriate transitions between WWW and DBS have to be applied. By means of a catalogue of demands, we will present the features of current methods, assign them to corresponding architectures of information processing and discuss the possibilities of further developments.

Semantic relationship and identification of music

Automatic search engines for musical documents mostly provide a vast number of relevant and irrelevant hits including multiple appearances of the same documents. Generally, underlying information-processing methods that are based on metadata or simple content-based procedures cause this situation. Therefore we propose a model for the recognition of music that enables identification and comparison of musical documents regardless their actual instantiation. Inherent component of this model is the capability of navigation based on semantic relationship between musical pieces. In order to realize this functionality we introduce a concept that tries to derive deductive music recognition from generic music production as an equivalent counterpart. Hence, some deductive processes of human music perception are simulated. The resulting structures enclose musical meaning and therefore can be used for the estimation of identity and relationship between musical documents. As a byproduct of this functionality, plagiarism and copyright infringements could be detected.

Towards Content-Related Indexing in Databases

Modern business appliications require huge volumes of highdimensioinal data to be Stored. explorative queries, typically used in these applications, select groups of objects with similar attributes or attribute combinations. In contrast to multidimensional index structures designed for spatial data that assume dimension independence and very often a uniform distribution, we have developed a new database indexing concept that discovers correlation patterns and takes the nonuniform distribution into consideration. The corresponding analysis is done on the subsymbolic level by applying a hierarchical artificial neural network. The trained neural network organises the data into a hierarchy of clusters. The clusters can be interpreted as groups of similar objects on the symbolic level. The hierarchy is finally used to derive the Intelligent Cluster Index (ICIx). In this paper we present a description of the Intelligent Cluster Index, it’s creation and application as multidimensional index and as heuristic for a logical distribution schema. We describe first experimental results, showing that this new approach can significantly speed up the system performance.

Deriving Composite Periodic Patterns from Database Audit Trails

Information about the periodic changes of intensity and structure of database workloads plays an important role in performance tuning of functional components of database systems. Discovering the patterns in workload information such as audit trails, traces of user applications, sequences of dynamic performance views, etc. is a complex and time consuming task. This work investigates a new approach to analysis of information included in the database audit trails. In particular, it describes the transformations of information included in the audit trails into a format that can be used for discovering the periodic patterns in database workloads. It presents an algorithm thatthe fluctuations finds elementary periodic patterns through nested iterations over a four dimensional space of execution plans of SQL statements and positional parameters of the patterns. Finally, it shows the composition rules for the derivations of complex periodic patterns from the elementary and other complex patterns.

Mining Periodic Patterns from Nested Event Logs

Information about periodic computations of processes, events, and software components can be used to improve performance of software systems. This work investigates mining periodic patterns of events from historical information related to processes, events, and software components. We introduce a concept of a nested event log that generalizes historical information stored in the application traces, event logs and dynamic profiles. We show how a nested event log can be compressed into a reduced event table and later on converted into a workload histogram suitable for mining periodic patterns of events. The paper defines a concept of periodic pattern and its validation in a workload histogram. We propose two algorithms for mining periodic patterns and we define the quality indicators for the patterns found. We show, that a system of operations on periodic patterns introduced in this work can be used to derive new periodic patterns with some of the quality indicators better from the original ones. The paper is concluded with an algorithm for deriving periodic patterns with the given quality constraints.

Profit of extending standard relational databases with the Intelligent Cluster Index (ICIx)

In this paper, we present a strategy to reduce the processing time needed for selection operations with many attributes in standard database systems. These problems mostly occur in data mining, data analysis, information retrieval, and applications with high combinatorial complexity. In these systems, standard indexes do not gain a satisfying performance. Currently, this problem is tackled using more computing power or special solutions instead of standard databases. Our approach is to interpret the queries as high-dimensional point or range queries. Thus, we provide a “real” solution to answering complex queries rather than merely postponing the problems using technical methods. We show the benefit of multidimensional data structures. This benefit can be transferred to a lot of applications (e.g. Business Intelligence, Bill Of Materials Explosion, DNA Sequence Search), not only advanced applications of database systems like GIS, CAD, or multimedia. Finally, a very small sample data set is used to show the profit of our approach and we present possible integration points of the ICIx into standard relational database.

On the Query Treatment in Federated Systems

In this paper, a systematic method for evaluating queries issued to an federated relational database system is presented. The method consists of four phases: syntactic analysis, query decomposition, query translation and result synthesis, which all can be done automatically based on the metadata built in the system, even if the structure conflicts among the local databases exist.

MapReduce-Based Growing Neural Gas for Scalable Cluster Environments

Growing Neural Gas (GNG) constitutes a neural network algorithm to create topology preserving representations of data, thus, being applicable in cluster analysis. With fast growing amounts of data, cluster analysis tasks face distributed data sets managed by cluster environments requiring scalable, parallel computation methods. In this paper we present a MapReduce-based version of the GNG training method. The algorithm is able to process large data sets on scalable cluster systems. We discuss its complexity and consider communication costs that arise from its structure. We conduct experiments on artificial data in different cluster environments to evaluate the algorithms scalability. Finally, we show that the algorithm is applicable for cluster analysis of large data sets in scalable cluster systems.

Predicting database workloads through mining periodic patterns in database audit trails

Information about the periodic changes of intensity and structure of database workloads plays an important role in performance tuning of functional components of database systems. Discovering the patterns in workload information, such as audit trails, traces of user applications, and sequences of dynamic performance views, is a complex and time-consuming task. This work investigates a new approach to analysis of information included in the database audit trails. In particular, it describes the transformations of information included in the audit trails into a format that can be used for discovering the periodic patterns in the fluctuations of database workloads. It presents an algorithm that finds elementary periodic patterns through nested iterations over a four-dimensional space of execution plans of SQL statements and positional parameters of the patterns. It proposes a collection of composition rules for the derivations of complex periodic patterns from the elementary and other complex patterns and it shows how to use such rules to predict the future workload levels.