Athens University Economics and Business (AUEB)
A recommender system for E-Tourism services
Abstract
dc:descriptionThis thesis focuses on the provision new, innovative ideas on recommendation using text analysis methods and complicated decision making process using fuzzy logic techniques in recommendation system filed. 2.779 reviews were collected by 43 London hotels which were processed by KNIME text analysis tool. This process calculated the significance of each service mentioned by each user. The profile of each user is formed by applying AHP in the significance service matrix. The hotel profile is built based the sum of the multiplication of customer preferences, the customer sentiment as expressed in the reviews and the time a review was written normalized by the total number of reviews written for the hotel . In terms representing hotel services is assigned a fuzzy number for each service mentioned by the user thus making better evaluation of the hotel services. Also the consideration of time in recommendation process allows for the changes in hotel services quality to be captured and to be taken into consideration when recommendations are produced. For the evaluation of the effectiveness of the algorithm the preferences of 207 hotel customers were collected through questionnaire and were multiplied by hotel profiles formed earlier. Thus the recommendation score of each hotel for each user was calculated. The effectiveness of the algorithm proposed is 73,4%.
Degree
thesis:*- Grantor dc:publisher
- Athens University Economics and Business (AUEB)
- Year dc:date
- 2019
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Tsetseka, Athina
- Τσέτσεκα, Αθηνά
Subjects
dc:subject × 9Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/46370