ETD

Archivio digitale delle tesi discusse presso l'Università di Pisa

Tesi etd-04272015-143002


Tipo di tesi
Tesi di dottorato di ricerca
Autore
GAZZE', DAVIDE
URN
etd-04272015-143002
Titolo
Social Media Monitoring and Analysis: Multi-domain Perspectives
Settore scientifico disciplinare
ING-INF/05
Corso di studi
INGEGNERIA "L. DA VINCI"
Relatori
relatore Ing. Tesconi, Maurizio
relatore Dott. Marchetti, Andrea
tutor Dott. Bechini, Alessio
Parole chiave
  • Web API
  • Social Media Analytics
  • crawling
  • Social Media
  • Social Network
  • capture
Data inizio appello
09/05/2015
Consultabilità
Completa
Riassunto
Social Media Platforms, such as Facebook or Twitter, are part of everyday life as powerful communication tools.
They let users communicate anywhere-anytime, improve their own public image and readily share information. For this reason, a growing number of individuals such as professionals as well as companies have opened an account in one or more Social Media platforms.
Due to the widespread use and growing numbers of users, a huge amount of data is generated every day. This information may play a crucial role in various decision-making processes.
In this setting, research topics connected to monitoring and analysis of Social Media data are becoming increasingly important.
The present work stems from data collection methodologies from different Social Media sources. It introduces the problems involved in storing semi-structured data, and in possible information gaps due to privacy policies. These facets are described according to the application domain as well as the Social Media platform. Subsequently, a theoretical generic architecture for handling data from Social Media sources is presented. We present three simplified versions of this architecture in three different domains: Online Reputation, Social Media Intelligence, and Opinion Mining in tourism. In the last part of the work, we introduce Social Media Analysis in these three domains. For the first, we present the project SocialTrends, a web application able to monitor “public” people on Facebook, Twitter, and YouTube.
In the second, we introduce an innovative approach for measuring the interactions between users in public spaces such as Facebook (public-by-design). Finally, we present Tour-pedia, a web application that displays a sentiment map of tourist locations in several cities according to different categories (accommodation, restaurants, points of interest and attractions).
File