Αλγόριθμοι Εξόρυξης Δεδομένων Με Εφαρμογές Στον Ιστό Και Σε Χρονικές Βάσεις Δεδομένων

Abstract

This thesis is focused in the field of Computer Science named Data Mining. Our work is based on one unpublished paper and two published ones. In the first unpublished one [64] we present a series of Power-Laws which we observed while working with Web Directories, for example in DMOZ. This work is extensively presented in chapter 3. Our chronologically first published paper [8] in SIAM Data Mining Conference concerns an innovative algorithm for the extraction of on-line communities within the Web. This algorithm is presented in chapter 2. Our second paper [7], published in the IEEE Conference on Data Mining, isabout a new algorithm we propose for searching within Temporal Databases of time series, while allowing several transformations in both axis. This work is presented in chapter 4. Apart from the above, in chapter 1 we present a short introduction in the field of Data Mining from both the algorithmic and applications aspect.