A study on tourism monitoring using social media and nighttime light remote sensing data | |
| Author | Devkota, Bidur |
| Call Number | AIT Diss no.RS-20-06 |
| Subject(s) | Tourism--Remote sensing Geographic information systems |
| Note | A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Remote Sensing and Geographic Information Systems |
| Publisher | Asian Institute of Technology |
| Abstract | The unprecedented popularity of novel data sources like Volunteered Geographic Information (VGI) and remotely sensed data provide a promising opportunity for understanding tourism conditions of a place. As the Internet is enriched with more user generated geo-tagged information, these publicly available data can be used to study various aspects of tourism. Petabytes of freely available satellite captured remote sensed data provides spatio-temporal clues regarding the socioeconomic conditions on earth. This research aims to explore the feasibility of collecting geodata from social network services (such as Twitter, Flickr), mapping service (i.e. OpenStreetMap) and nighttime light remotely sensed images to discover and describe tourism hot spots. Existing studies have applied various approaches to gather, analyze, identify and describe tourism places of interest based on such information. Such unconventional data sources offer easy, economical, and near-real-time identification of tourism areas of interest. This study demonstrates the potential of volunteered geographic information, mainly Twitter and OpenStreetMap for discovering tourism areas of interest. Active tweet clusters generated using Density-Based Spatial Clustering of Applications with Noise (DBSCAN) clustering algorithm and building footprint information are used to identify touristic places that ensure the availability of basic essential facilities for travelers. Furthermore, an investigation is made to examine the usefulness of nighttime light (NTL) remotely sensed data to recognize such tourism areas. The study successfully discovered important tourism areas in urban and remote regions in Nepal which have relatively low social media penetration. The effectiveness of the proposed framework was established with standard performance measure. The accuracy assessment showed F1 score of 0.72 and 0.74 in the selected regions. Further, machine learning algorithms were used to automate the process of Tourism Areas of Interest (TAOI) identification. Additional data from different sources such as OpenStreetMap, WorldPop, etc. were provided for improved model design. Experimental results demonstrated Random Forest (RF) algorithm as a better performer among the algorithms examined. Support Vector Machine (SVM) and RF classifiers obtained the best performance with F1 score of 0.82 based on the training and validation data from the same geographic region. Whereas, RF yielded the utmost score, i.e. 0.76, when the validation is done for new region. Hence, such machine learning models could effectively utilize the learning from a set of data to new data and even to another geographic region. This indicates the possibilities for universalization of the proposed method. Apart from identifying important TAOIs, it is important to know why they are important for the visitors. An approach, Term Frequency and Inverse Document Frequency (TFIDF) and word freshness based scoring scheme, is devised to determine representative keywords of tourism spots based on user experience embedded in social media text. The extracted keywords proved relevant during comparison with tourism expert's suggestion as well as real visitor's opinion. Comparison against independent data source using Mann-Whitney U test confirmed the similarity of the important keywords at a significance level of 0.05. Hence, this study can provide a valuable reference for more advanced studies on tourism, VGI and NTL. Also, various tourism stakeholders may find it relevant during tourism planning and management. |
| Year | 2020 |
| Type | Dissertation |
| School | School of Engineering and Technology |
| Department | Department of Information and Communications Technologies (DICT) |
| Academic Program/FoS | Remote Sensing and Geographic Information Systems (RS) |
| Chairperson(s) | Miyazaki, Hiroyuki |
| Examination Committee(s) | Apichon Witayangkurn;Kim, Sohee Minsun |
| Scholarship Donor(s) | Japanese Government |
| Degree | Thesis (Ph.D.) - Asian Institute of Technology, 2020 |