A city-scale event (CSE) is a large collection of small events that are centered around a specific theme, time and place, for examples EXPOs or conventions. These macro-events attract very large crowds and are particularly hard to manage and analyse, as they involve a number of different entities - from CSE/event managers to visitors. A crucial point in the analysis of a CSE is to understand which of the many events, venues and districts aroused most interest. Nowadays, this analysis can be done by volunteers that make surveys and collect statistics but this is a very expensive and time-consuming method. Since these data are vital in order to make the event more successful and profitable, we wondered if the data coming from social media could offer a valid and cheaper alternative. In our project we used Twitter as a source of user-generated information. Given a database with all the description of events/venues/districts we tried to find out what a certain Tweet is talking about. We defined a set of text-processing functions and similarity metrics to compare the tweets contents and hashtags with the names of events and venues in order to find if a given venue/event is mentioned inside the Tweet. We used the Silk Framework to do all these operations. We focused our attention on the Fuorisalone 2013 edition, a one week-long collection of design-related events spread throughout Milan that hosted more than 600 venues, 1200 events and drew about half a million visitors. We developed a multi-level application (CSE LiVeTweet Analyser) that offers five different detail levels in order to meet the needs of different users’ profiles, from the CSE organizer to the event manager. The results obtained show that CSE LiVeTweet Analyser offers a complementary solution to the traditional text-based search as it relies on the relative importance of the ’meaningful word(s)’ in the context of the whole tweet. Our method generates with high level of precision (70%), about a half of the links found by the text-based search. In addition CSE LiVeTweet Analyser generates one third of new links, thanks to the contemporary search of venue’s and event’s information.

CSE LiVeTweet analyser : a system for generating links between Venues and Tweets during city scale events

RE CALEGARI, GLORIA;NASI, GIOELE
2013/2014

Abstract

A city-scale event (CSE) is a large collection of small events that are centered around a specific theme, time and place, for examples EXPOs or conventions. These macro-events attract very large crowds and are particularly hard to manage and analyse, as they involve a number of different entities - from CSE/event managers to visitors. A crucial point in the analysis of a CSE is to understand which of the many events, venues and districts aroused most interest. Nowadays, this analysis can be done by volunteers that make surveys and collect statistics but this is a very expensive and time-consuming method. Since these data are vital in order to make the event more successful and profitable, we wondered if the data coming from social media could offer a valid and cheaper alternative. In our project we used Twitter as a source of user-generated information. Given a database with all the description of events/venues/districts we tried to find out what a certain Tweet is talking about. We defined a set of text-processing functions and similarity metrics to compare the tweets contents and hashtags with the names of events and venues in order to find if a given venue/event is mentioned inside the Tweet. We used the Silk Framework to do all these operations. We focused our attention on the Fuorisalone 2013 edition, a one week-long collection of design-related events spread throughout Milan that hosted more than 600 venues, 1200 events and drew about half a million visitors. We developed a multi-level application (CSE LiVeTweet Analyser) that offers five different detail levels in order to meet the needs of different users’ profiles, from the CSE organizer to the event manager. The results obtained show that CSE LiVeTweet Analyser offers a complementary solution to the traditional text-based search as it relies on the relative importance of the ’meaningful word(s)’ in the context of the whole tweet. Our method generates with high level of precision (70%), about a half of the links found by the text-based search. In addition CSE LiVeTweet Analyser generates one third of new links, thanks to the contemporary search of venue’s and event’s information.
BALDUINI, MARCO
ING - Scuola di Ingegneria Industriale e dell'Informazione
28-apr-2014
2013/2014
Tesi di laurea Magistrale
File allegati
File Dimensione Formato  
2014_04_ReCalegari_Nasi.pdf

solo utenti autorizzati dal 09/04/2015

Descrizione: testo della tesi
Dimensione 2.23 MB
Formato Adobe PDF
2.23 MB Adobe PDF   Visualizza/Apri

I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10589/89962