EBV: Encoded Binary Vector for Efficient Information Retrieval, Query Processing and Recommendation for Travel and Touri
- PDF / 1,412,238 Bytes
- 16 Pages / 595.276 x 790.866 pts Page_size
- 14 Downloads / 169 Views
RESEARCH ARTICLE-COMPUTER ENGINEERING AND COMPUTER SCIENCE
EBV: Encoded Binary Vector for Efficient Information Retrieval, Query Processing and Recommendation for Travel and Tourism Domain EBV for Travel and Tourism Recommendation Jobi Vijay1
· Rajeswari Sridhar2
Received: 29 August 2019 / Accepted: 23 September 2020 © King Fahd University of Petroleum & Minerals 2020
Abstract The rise in business and electronic transactions using the web as a medium has driven the widespread use of recommendation systems across all domains. Recommendation systems generate suggestions for the travel and tourism industry, particularly hotels or services, centered on user interests and preferences. Relevant information collection, organization, processing vast quantities of string data, efficient data retrieval, and computational complexity in handling information overload are key issues to be addressed in a recommendation system. This work proposes a novel approach, using encoded bit vectors, to process information on hotel amenities and features extracted from Tripadvisor. Hotel amenities at different locations are organized and stored, based on the priority of the location, using the most significant bit and least significant bit. The amenities are prioritized dynamically thereafter, based on the location, and unique amenities accorded greater preference than the ones routinely offered at most hotels. Encoded bit vectors are utilized for information retrieval and precise hotel recommendations made, based on user interest rather than the string processing method that is ineffective where scalable data are concerned. Finally, the recommended hotels are ranked with user-centric personalized recommendations. These highspeed, efficient, and scalable algorithms are ideal for incrementally updating data. They also facilitate fast, flawless IR and query processing, along with appropriate hotel recommendations. The theoretical evaluation shows that the proposed method aids quick query processing and retrieval. The experimental assessment shows that hotels recommended by the proposed bit vector recommendation system have a higher accuracy on par with user and expert recommendations. Keywords Information retrieval (IR) · Encoded bit vector · Most significant bit (MSB) · Least significant bit (LSB) · Tripadvisor · Recommendation system
1 Introduction User-generated social media data are increasing exponentially on a daily basis, calling for effective IR and recommendation systems that analyze user interests to locate ideal products for their recommendation [1]. This is unchartered territory, considering the number of efforts directed toward
B
Jobi Vijay [email protected] Rajeswari Sridhar [email protected]
1
Department of Computer Science and Engineering, Anna University, Chennai, India
2
Department of Computer Science and Engineering, National Institute of Technology, Trichy, India
developing an appropriate web search paradigm. There is always a need for competent information collection, preprocessing, indexing, storage,
Data Loading...
