|
Title:
|
A PROBABILISTIC TOPIC MODEL-BASED RECOMMENDATION APPROACH FOR SIGHTSEEING SPOTS |
|
Author(s):
|
Satona Shirai, Hiroshi Uehara and Kazuya Suzuki |
|
ISBN:
|
978-989-8704-36-8 |
|
Editors:
|
Piet Kommers, Tomayess Issa, Adriana Backx Noronha Viana, Theodora Issa and Pedro Isaías |
|
Year:
|
2021 |
|
Edition:
|
Single |
|
Keywords:
|
Data Mining, Data Analytics, Recommendation, Tourism, Machine Learning |
|
Type:
|
Poster |
|
First Page:
|
225 |
|
Last Page:
|
227 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
In this paper, we propose a recommendation approach for sightseeing spots using travelers' reviews. The task is
predicting suitable sightseeing spots for newly posted travelers' reviews based on the datasets generated in our previous
study (Shirai et al., 2021), which provides foreign visitors with a wide variety of reviews accompanied by
recommendable sightseeing spots. By applying joint topic model, a Bayesian inference method, both the reviews in the
augmented dataset and newly posted reviews are represented as probabilistic distribution of preferences, thereby very
similar reviews in the augmented dataset to each newly posted one are detected. We evaluated accompanying sightseeing
spots to these very similar reviews in terms of how these spots are coincident with the one attached to the newly posted
review. The result showed several reviews at the top of the similarity ranking were accompanied by the sightseeing spots
coincident with the one of the newly posted review. |
|
|
|
|
|
|
