ระบบจัดเก็บและค้นหางานวิจัยอัจฉริยะ

Prediction of Tourist Behaviour: Tourist Visiting Places by Adapting Convolutional Long Short-Term Deep Learning

จำนวนผู้เข้าชม

Prediction of Tourist Behaviour: Tourist Visiting Places by Adapting Convolutional Long Short-Term Deep Learning

ดาวน์โหลด: 0 ครั้ง

Collection	RMUTK Research Repository (RMUTK IR) บทความประชุมวิชาการ — Proceedings
ID RMUTK Digital	RMUTK000128
Title	Prediction of Tourist Behaviour: Tourist Visiting Places by Adapting Convolutional Long Short-Term Deep Learning
Alternative title	Prediction of Tourist Behaviour: Tourist Visiting Places by Adapting Convolutional Long Short-Term Deep Learning
Authors	ผศ.ดร.จารุวรรณ กาญจนศุภวรรณ ผู้แต่งหลัก
Faculty	คณะบริหารธุรกิจ
Volume / Issue / Pages	pp.12-17
Conference	2019 International Conference on System Science and Engineering (ICSSE) (The Best Paper Award)
Conference Date	2019-07-20
Conference Location	Dong Hoi, Vietnam
Year	2562
Level	ระดับนานาชาติ
DOI	10.1109/ICSSE.2019.8823542
ISBN / ISSN	978-1-7281-0525-3
Funding Source	-
Abstract	-
Abstract (EN)	The trend of industry tourism GDP is increasing in every year that speculates from statistics of the World Travel & Tourism Council (2018). Moreover, travel industry not only considered as the most dynamic sector but also the most importance generator of income and jobs in the country. Thus, the prototype for tourism plans are needed for strategic planning. Currently, social web is a great tool for providing useful insights about tourist behaviors especially with the text data that comes from travelers' opinions. In this work, we use sequential patterns of users' behavior which are ordered by time from tourist including opinions, reviews as our input data. Then, we use Convolutional Long Short-Term Deep Learning (CLSTDL) which is a deep learning technique that combines Convolutional Neural Network (CNN) with Long Short-Term Memory (LSTM) to predict the expected location. During the process, the output of CNN will be fed into LSTM to learn the sequence pattern behavior of traveler. The model output is then used to predict the next location that particular travelers are likely to go. The experimental results have shown that CLSTDL outperforms other models when evaluating with the accuracy and loss metrics.
Keywords (TH)	-
Keywords (EN)	LSTM CNN sequence learning Convolutional Long Short-Term Deep Learning CLSTDL
Access Level	Open Access (เปิดสาธารณะ)

เอกสารดิจิทัล

RMUTK000128_full.pdf

จำนวนดาวน์โหลด: 0 ครั้ง

Open Access

ดาวน์โหลด