Browsing by Author "Darcan, Osman Nuri."

Now showing 1 - 7 of 7
  • Thumbnail Image
    Item
    A Java based development environment for SpringFramework
    (Thesis (M.S.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2008., 2008.) Kasap, Fatih.; Darcan, Osman Nuri.
    Web application development is one of the most important programming areas today. There are many technologies to provide a better and easier way of programming for web applications in the market. Some of them are open source and free whereas some of them are released by big software vendors such as Microsoft and IBM. The SpringFramework is an open source alternative in this area. It is a Java-based technology to provide a robust infrastructure to software development based on best practices and accepted standards. This thesis aims to provide a computer program to facilitate SpringFramework-based web application development by hiding low level tasks from the developer and trying to make it focus on the real purpose of his application. Based on user friendliness and being close to human nature this program provides automatic code and web page generation, database and internationalization support, easy file management.
  • Thumbnail Image
    Item
    A study on sequential internet auctions using agent-based modeling approach
    (Thesis (M.A.)-Bogazici University. Institute for Graduate Studies in Social Sciences, 2009., 2009.) Akkaya, Yıldız.; Badur, Bertan Yılmaz.; Darcan, Osman Nuri.
    With widespread use of the Internet, Internet auctions (e-auctions) become more popular in order to trade increasing number of goods as Internet provides both almost perfect market information and an infrastructure for executing auctions at lower administrative costs. The sequential auctions are the most widely used auction format. The aim of this study is to present a dynamic model of an e-auction so as to investigate how the welfare of buyers is affected by different bidding strategies. This problem has been studied in economics by conducting laboratory and field experiments and theoretically in various static auction mechanisms where perfect rationality of participants is assumed. On the other hand, observing the biding strategies of individuals is almost impossible in laboratory or field experiments. To overcome the limitations of these approaches, the new agent-based modeling methodology in which researchers use simulations to investigate the behavior and interactions of autonomous, heterogeneous, boundedly rational adaptive population of agents in the social and economical environments, has been emerged. In the study, the bottom-up agent-based modeling and simulation methodology is adapted to investigate the behavior of participants in electronic markets. A simulation model is developed to understand the effects of different bidding and bid increment strategies on the welfare of the bidder. To some extend sensitivity of the auction outcome on auction rules and market design parameters are also investigated.
  • Thumbnail Image
    Item
    An association rule-based recommendation engine for online dating sites
    (Thesis (M.A.)-Bogazici University. Institute for Graduate Studies in Social Sciences, 2009., 2009.) Özseyhan, Yağız Civan.; Badur, Bertan Yılmaz.; Darcan, Osman Nuri.
    In this study, the database of a Turkish online dating site is analyzed to reveal patterns in the personal features of matching couples. By using a functionality of data mining, the Association Rule mining, a set of rules is extracted from the available messaging and user data. The rules are used for the development of a recommendation engine which is capable to give list of best potential matches to the site user. The performance of the recommendation engine is tested with statistical tools to find whether the increase in the quality of results is significant.
  • Thumbnail Image
    Item
    An intelligent database interface for Turkish
    (Thesis (M.S.)- Bogazici University. Institute for Graduate Studies in Sciences and Engineering, 1991., 1991.) Darcan, Osman Nuri.; Kuru, Selahattin,
    In this thesis, a portable natural language interface system for communicating with databases in Turkish is developed. The system does a two step transformation from a Turkish query in user's view to an intermediate meaning repl'esentation language D&Q and finally to a target database language SQL. It is composed of domain independent runtime modules for different processing stages, namely language processing, internal query generation and translation to SQL. Modules refer to the knowledge base in which diverse knowledge about the domain and the database are maintained. Two additional modules, to wita spelling corrector and a history keeper are incorporated in the system. A syntactic parser is used in analyzing queries. for the syntactic parser. a formalization of a subset of Turkish grammar based on the simple principle of general categorization incorporated with the notion of modifications between words is proposed and a grammar that consists of a collection of rervrite rllles for the formal representation of sentences is discussed. A decision tree which works with suffix strip off approach is used for the morphological parsing of n.ouns. Parse trees produced for different types of sentences using the formalized grammar are given. Regarding to the meaning representation, an "intelllgent" meaning representation generator which has a rule based reasoning capability is designed. The interpretations of some modification relations are discussed in details. Finally the interpretation of a full sentence is shown. The system is tested on an imaginary stude-course-instructor database, all examples refers to this database. Possible extensions for both the parser and the meaning representation generator are also proposed at the end of the thesis.
  • Thumbnail Image
    Item
    Distributed simulation framework based on load balanced implementation of standard clock algorithm with web based extensions
    (Thesis (Ph.D.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2003., 2003.) Darcan, Osman Nuri.; Kaylan, Ali Rıza.
    Standard Clock approach is used to simulate a number of parametric variants of a single system. In this thesis, a simulation engine based on the distributed implementation of the Standard Clock approach on networks of heterogeneous UNIX workstations is developed. The objective is to examine the scalability of the implementation and study the effect of load balancing. Two different heuristic load balancing techniques are proposed: (1) a static load balancing that is based on estimated cost of each variant and (2) a dynamic load balancing that migrates variants between workstations, based on their estimated performance during the simulation process. Simple queueing models are used to study the performance. Numerical results obtained from real-time simulations on a network of up to 7 workstations are used to investigate the speedup and the efficiency of both the implementation and the load balancing techniques. As more workstations are added to the simulation environment, a sublinear speedup is obtained. In addition, the load balanced distribution has produced an improvement up to 8 per cent compared to random distribution. Secondly, a web based user interface is integrated to this engine to provide an easy to use and practical experimentation platform. The complete system mainly consists of a web based graphical user interface that communicates with a powerful server that runs the engine. The user interface allows creating simulation models of queueing network, performing simulation experiments using the simulation engine, and performing simple output analysis in a platform independent manner. Finally, a prototype for a web based simulation optimization tool is developed by extending the system with the power of experimental design methodology and response surface method. Hence, this tool provides a way of planning which variants to simulate in order to quickly reach the neighbor of the optimal solution. The applicability of this simulation optimization approach and the features of the interface are illustrated using a two-node Jackson network and a manufacturing system as examples.
  • Thumbnail Image
    Item
    Empirical and simulation based analysis of detailing activities of a pharmaceutical company in Turkey
    (Thesis (M.A.) - Bogazici University. Institute for Graduate Studies in the Social Sciences, 2014., 2014.) Demirbacak, Nuri Sezay.; Darcan, Osman Nuri.
    In this thesis, the impact of detailing activities on prescription behaviour of physicians and their carryover effects are studied using the geographic location based sales and detailing activity data of five products from a leading pharmaceutical company in Turkey. First, the relationship of sales to detailing activities and to activities made in different customer segments are analyzed using a linear regression model. It has been found that a significant relationship exists between total detailing activities and sales for four out of five products. As to the activities made in different customer segments, significant results are observed for only one product. Secondly, the outcomes of the regression are used to develop an agent based model to predict the impact and the carryover effect of detailing activities on different customer segments. The outcomes of the simulation model, when run with the best fitted impact and carryover ratios for the selected regions, reveal 10-15% deviation amongst the actual monthly values and the predicted values.
  • Thumbnail Image
    Item
    Experimental study for extending data mining standards
    (Thesis (M.A.)-Bogazici University. Institute for Graduate Studies in the Social Sciences, 2008., 2008.) Daylan, Ahmet.; Badur, Bertan Yılmaz.; Darcan, Osman Nuri.
    Data mining is becoming a mainstream technology used in business intelligence solutions supporting various industries and lines of business. Although there are plenty of data mining products at the market, these products are difficult to integrate with user applications due to the lack of standardization protocols. Conforming to common standards facilitates development, implementation and maintenance of applications as well as communication among them. In addition, these standards enable data miners to develop data mining process easily. As being one of the wellestablished data mining standards, Java Data Mining (JDM) allows Java applications to communicate with data mining engines to build, test and apply mining models. First release of JDM supports the basic data mining functionalities; classification, regression, attribute importance, clustering, and association. Currently developing version (JDM 2.0) proposes the additional set of functionalities such as time series, feature extraction, text mining. In this study, we present the experimental framework developed to extend the JDM 2.0 by including ensembles methods such as boosting, bagging to improve the classification accuracy and bootstrap to assess accuracy. We have applied our extended JDM functionalities with two well known datasets that are iris dataset and Sales History (SH) dataset.