ETD system

Electronic theses and dissertations repository

 

Tesi etd-09232008-165307


Thesis type
Tesi di laurea specialistica
Author
BACCIANELLA, STEFANO
URN
etd-09232008-165307
Title
Multi-facet rating of online hotel reviews: issues, methods and experiments
Struttura
SCIENZE MATEMATICHE, FISICHE E NATURALI
Corso di studi
INFORMATICA
Commissione
Relatore Dott. Sebastiani, Fabrizio
Relatore Dott. Esuli, Andrea
Relatore Prof. Ferragina, Paolo
Parole chiave
  • reviews
  • regression
  • pattern
  • tripadvisor
  • rating
  • hotel
  • jatecs
  • classifier
Data inizio appello
10/10/2008;
Consultabilità
completa
Riassunto analitico
Online product reviews are becoming increasingly popular, and are being<br>used more and more frequently by consumers in order to choose among<br>competing products. Tools that rank competing products in terms of the<br>satisfaction of consumers that have purchased the product before, are thus<br>also becoming popular. We tackle the problem of rating (i.e., attributing<br>a numerical score of satisfaction to) consumer reviews based on their tex-<br>tual content. In this work we focus on multi-facet rating of hotel reviews,<br>i.e., on the case in which the review of a hotel must be rated several times,<br>according to several aspects (e.g., cleanliness, dining facilities, centrality of<br>location). We explore several aspects of the problem, including the vectorial<br>representation of the text based on sentiment analysis, collocation analysis,<br>and feature selection for ordinal-regression learning. We present the results<br>of experiments conducted on a corpus of approximately 15,000 hotel reviews<br>that we have crawled from a popular hotel review site.
File