Back to search

National Technical University of Athens (NTUA)

Εξατομικευμένα συστήματα διαχείρισης δεδομένων

Abstract

dc:description

The exponential growth of the Web has turned it into a huge repository of information for humanity. Key features of this development are: (a) the oversupply of information and services, (b) the coverage of the demands for users with various interests and characteristics, and (c) the variety of information accessing means. Furthermore, the evolution of social networks and Web 2.0, and the growing internet access via fast wireless networks and mobile devices (e.g. smartphones, tablets) have enabled the collection of vast amounts of data regarding the characteristics, interests and preferences of users and the corresponding usage scenarios. In this context, several challenges emerge for both users and information providers, since for the former the task of discovering the portions of information that are the most relevant to their interests and preferences is becoming harder, while the latter seek ways for more effective and targeted promotion of their services to the users who are more likely to be interested in them. Hence, the need for systems and methods that provide personalized services to users and providers is becoming urgent. This thesis deals with personalization issues in data management at both system and application levels. Specifically, we focus on: (a) user preference management for relational data, and (b) preference query evaluation techniques. For the first part, we propose a new model for representing preferences over relational data and we extend the existing relational algebra operators, such that they take into account the user preferences. Moreover, we have implemented a prototype preference-aware data management framework embedded in a standard relational database system. With regards to preference queries evaluation we approach the problem: (a) from a user's perspective where we propose algorithms for processing skyline queries when user preferences depend on the usage context, e.g., location, time, weather, social environment, and (b) from the view of data providers, where we propose methods for efficient processing of market research queries. In particular, we propose algorithms: (a) for identifying those users that are more likely to buy a product, with applications in personalized advertising, and (b) for optimal product positioning, where we aim to maximize the profit measured as the estimated number of potential product buyers.

Degree

thesis:*
Grantor dc:publisher
National Technical University of Athens (NTUA)
Year dc:date
2013

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Arvanitis, Anastasios
  • Αρβανίτης, Αναστάσιος

Subjects

dc:subject × 19

Rights

Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/29115

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Arvanitis, Anastasios; Αρβανίτης, Αναστάσιος. Εξατομικευμένα συστήματα διαχείρισης δεδομένων. National Technical University of Athens (NTUA), 2013. http://hdl.handle.net/10442/hedi/29115