logo SBA

ETD

Archivio digitale delle tesi discusse presso l’Università di Pisa

Tesi etd-02072017-201541


Tipo di tesi
Tesi di laurea magistrale
Autore
DE BIANCHI, LUIGI
URN
etd-02072017-201541
Titolo
Design and implementation of a system based on computational stigmergy to identify event-specific social discussion dynamics
Dipartimento
INGEGNERIA DELL'INFORMAZIONE
Corso di studi
COMPUTER ENGINEERING
Relatori
relatore Prof. Cimino, Mario Giovanni Cosimo Antonio
relatore Prof.ssa Vaglini, Gigliola
Parole chiave
  • clustering
  • computational stigmergy
  • data gathering
  • differential evolution
  • genetic algorithm
  • som
  • twitter
Data inizio appello
24/02/2017
Consultabilità
Completa
Riassunto
Social networks are a mine of information and various approaches were used to extract meaningful data particularly from Twitter; it is the most common medium to gather social user messages since offers public streams of messages. A lot of methods have been proposed in the literature to try to identify peaks of interest related to a specific term or to identify interactions between groups of related terms. The proposed approach, instead of following a group of terms, follows an event characterized by some terms and analyses the most common words using the stigmergic approach evaluating similarity in time and Self-Organizing Map to summarize the event exploiting the behaviour of each term. Tweets gathered during the third American presidential debate where analysed as proving ground for this method. The system was able to elaborate a coherent map able to condense the event and the different phases of the debate proving itself better than other similarity measures proposed in the literature.
File