|
|
Preference Top-k Search Based on Multidimensional B-tree
Ondreička, Matúš ; Pokorný, Jaroslav (advisor) ; Theobald, Martin (referee) ; Gurský, Peter (referee)
Title: Preference Top-k Search Based on Multidimensional B-Tree Author: RNDr. Matúš Ondreička Department: Department of Software Engineering Faculty of Mathematics and Physics Charles University in Prague Supervisor: Prof. RNDr. Jaroslav Pokorný, CSc. Author's e-mail address: ondreicka@ksi.mff.cuni.cz Supervisor's e-mail address: pokorny@ksi.mff.cuni.cz Abstract: In this thesis, we focus on the top-k search according to user pref- erences by using B+ -trees and the multidimensional B-tree (MDB-tree). We use model of user preferences based on fuzzy functions, which enable us to search according to a non-monotone ranking function. We propose model of sorted list based on the B+ -tree, which enables Fagin's algorithms to search for the top-k objects according to a non-monotone ranking function. We apply this model in the Internet environment with data on different remote servers. Furthermore, we designed novel dynamic tree-based data structures, namely, MDB-tree composed of B+ -trees, MDB-tree with lists, MDB-tree with groups of B+ -trees and multiple-ordered MDB-tree. Concurrently, we have developed novel top-k algorithms, namely, the MD algorithm, the MXT algorithm and their variants which are able search for the top-k objects ac- cording to a non-monotone ranking function. These top-k algorithms are efficient...
|
|
|
Preferencev querying, indexing, optimisation
Horničák, Erik ; Vojtáš, Peter (advisor) ; Ondreička, Matúš (referee)
In this thesis we discuss the issue of searching the best k objects from the multi-users point of view. Every user has his own preferences, which are represented by fuzzy functions and aggregation function. This thesis designs and implements several solutions of searching the best k objects when attributes data are stored on remote servers. It was necessary to modificate existing algorithms for this type of obtaining data. This thesis uses several variants of Fagin algorithm, indexing methods using B+ trees and communication via web services.
|
|
|
Detection of Plagiarism
Dufková, Kateřina ; Galamboš, Leo (advisor) ; Ondreička, Matúš (referee)
The master thesis analyses the possibilities of web search engine methods application to the problem of e ective plagiarism detection. It focuses on 1:N plagiarism detection methods under the circumstance that the corpus is a very large dynamically changing set of documents. For the implementation the probabilistic aproximation of the Jaccard similarity measure using hashing combined with shingling was chosen. The aim of the thesis is to present an imlementation of this method within the Egothor 2 web search engine, to describe the features and parameters of the implementation and nally to evaluate the advantages and contingent limitations of the approach.
|
|
|
Models of user preferences in e-shop environment
Václav, Branislav ; Vojtáš, Peter (advisor) ; Ondreička, Matúš (referee)
The aim of this work is to gain insight into the broad range of models of user preferences inside an e-shop environment. A specific group of models will be selected from the overall described set, and an exact method of calculation for these models will be introduced. The selected models, together with a corresponding web environment design, will then be implemented into a comprehensive form of a working web application. An integral part of the application is formed by the inclusion of an appropriate set of test data. Based on these data, practical experiments will be carried out, and consequent results will be considered in the assessment of the functionality of the provided application and its potential contribution for existing e-shop users. Acquired user feedback will then be used to identify further development opportunities of the implemented application.
|
|
|
Analysis of user behaviour on web pages
Žák, Vladimír ; Eckhardt, Alan (advisor) ; Ondreička, Matúš (referee)
Diploma thesis gathers knowledge about current methods of analysis of user behaviour on web pages and interpretation of his (her) behaviour as his (her) preference. Conventional approach to get information about user need is using explicit relevance methods, where user explicitly rates a web page or object within the page. It can be di cult to collect necessary data, so we consider the use of implicit feedback techniques. These techniques unobtrusively obtain information about user and remove the cost to the user of providing feedback. We describe some of these methods, then we create a measuring of effectivity of them. The work covers implementation of choosen methods and testing on a web to show practical value.
|
|
|
Extending Fagin's algorithm for more users
Ondreička, Matúš
We discuss the issue of searching the best K objects in more attributes for more users. Every user prefers objects in different ways. User preferences are modelled locally with a fuzzy function and globally with an aggregation function. Also, we discuss the issue of searching the best K objects without accessing all objects. We deal with the use of local preferences when computing Fagin's algorithms. We created a new model of lists for Fagin's algorithms based on B+-trees. Furthermore, we deal with the use of a multidimensional B-tree for searching the best K objects. We developed an MD-algorithm, which can effectively find the best K objects in a multidimensional B-tree in accordance with user's preferences and without accessing all the objects. Last but not least, we show results of all the tests of described algorithms. MD-algorithm achieves better results in the number of accessed objects than Fagin's algorithms.
|
|
|
Preference Top-k Search Based on Multidimensional B-tree
Ondreička, Matúš ; Pokorný, Jaroslav (advisor) ; Theobald, Martin (referee) ; Gurský, Peter (referee)
Title: Preference Top-k Search Based on Multidimensional B-Tree Author: RNDr. Matúš Ondreička Department: Department of Software Engineering Faculty of Mathematics and Physics Charles University in Prague Supervisor: Prof. RNDr. Jaroslav Pokorný, CSc. Author's e-mail address: ondreicka@ksi.mff.cuni.cz Supervisor's e-mail address: pokorny@ksi.mff.cuni.cz Abstract: In this thesis, we focus on the top-k search according to user pref- erences by using B+ -trees and the multidimensional B-tree (MDB-tree). We use model of user preferences based on fuzzy functions, which enable us to search according to a non-monotone ranking function. We propose model of sorted list based on the B+ -tree, which enables Fagin's algorithms to search for the top-k objects according to a non-monotone ranking function. We apply this model in the Internet environment with data on different remote servers. Furthermore, we designed novel dynamic tree-based data structures, namely, MDB-tree composed of B+ -trees, MDB-tree with lists, MDB-tree with groups of B+ -trees and multiple-ordered MDB-tree. Concurrently, we have developed novel top-k algorithms, namely, the MD algorithm, the MXT algorithm and their variants which are able search for the top-k objects ac- cording to a non-monotone ranking function. These top-k algorithms are efficient...
|
|
|
Analysis of user behaviour on web pages
Žák, Vladimír ; Eckhardt, Alan (advisor) ; Ondreička, Matúš (referee)
Diploma thesis gathers knowledge about current methods of analysis of user behaviour on web pages and interpretation of his (her) behaviour as his (her) preference. Conventional approach to get information about user need is using explicit relevance methods, where user explicitly rates a web page or object within the page. It can be di cult to collect necessary data, so we consider the use of implicit feedback techniques. These techniques unobtrusively obtain information about user and remove the cost to the user of providing feedback. We describe some of these methods, then we create a measuring of effectivity of them. The work covers implementation of choosen methods and testing on a web to show practical value.
|
|
|
Preferencev querying, indexing, optimisation
Horničák, Erik ; Vojtáš, Peter (advisor) ; Ondreička, Matúš (referee)
In this thesis we discuss the issue of searching the best k objects from the multi-users point of view. Every user has his own preferences, which are represented by fuzzy functions and aggregation function. This thesis designs and implements several solutions of searching the best k objects when attributes data are stored on remote servers. It was necessary to modificate existing algorithms for this type of obtaining data. This thesis uses several variants of Fagin algorithm, indexing methods using B+ trees and communication via web services.
|
|
|
Detection of Plagiarism
Dufková, Kateřina ; Ondreička, Matúš (referee) ; Galamboš, Leo (advisor)
The master thesis analyses the possibilities of web search engine methods application to the problem of e ective plagiarism detection. It focuses on 1:N plagiarism detection methods under the circumstance that the corpus is a very large dynamically changing set of documents. For the implementation the probabilistic aproximation of the Jaccard similarity measure using hashing combined with shingling was chosen. The aim of the thesis is to present an imlementation of this method within the Egothor 2 web search engine, to describe the features and parameters of the implementation and nally to evaluate the advantages and contingent limitations of the approach.
|
guest ::
