Summary of ‘COVID-19 Using Data Analysis using Mobile Services Data and Satellite Geospatial Data towards Sustainable Tourism Business’

Introduction

Recent Trends in Satellite Data Applications

The first topic, “Recent Trends in Satellite Data Applications” was presented by Mr. Onda Yasushi (Graduate School of System Design and Management, Keio University, Japan). First of all, before explaining the definition of satellite data, Mr. Onda showed an example of a photo captured by a satellite. It shows an oil tanker lined up in the ocean during COVID-19, captured by a satellite orbiting around the earth. Based on these data from the satellite images, the analysis process is able to connect and conduct by the different methods. Then, he explained the satellite data knowledge, there are many kinds of satellites, such as a radar satellite, but the color images cannot be gotten. The satellite that can capture color remote sensing will take pictures of the images. It can also capture the color not visible to the human eyes, such as infrared, etc. Next, he showed the recent applications in satellite data. Next, he showed the recent applications in satellite data. The first example is Heat Island Analysis done in New York City, USA. This study was done using a LANSAT satellite to capture infrared images. By capturing that data, they can figure out the temperature on the surface of a particular area. Thus, New York city could figure out the hotspot and the cool zone in certain areas and provide a countermeasure on supporting certain areas/buildings to have a cooling system in the building to make lives better for people living there.

The other example is the satellite data used as the investment information. A US company, ARESMETRICS, used high-resolution imagery data from satellites to understand many different types of investment information. For example, if an investor wants to invest in metal production, a satellite can observe the condition of a metal refinery. Then, they can predict the metal yields. Suppose an investor wants to invest in a car company. In that case, they can observe the condition of the car factory through satellite data and decide whether it is worth investing in based on the activity in that factory that is captured by the satellite data. And the following example was about the energy supply chain analysis. The example that was brought up, in this case, is observing oil trends in certain oil facilities that people can understand how much oil is produced. Another case study was about weather insurance. They have predicted the weather conditions by using satellites, how much rain is actually felled, and if it will cause disaster (e.g., drought). Then, an insurance company can decide on insurance for farmers with a reasonable price with reasonable compensation.

Another example is a global fishing watch. It is possible to monitor the global fishing activities by the GPS tracking the shipping vessels with satellite imagery. Then, they can compare the officers to investigate their fishing activities if it has a suspicious intention, which is probably doing illegal fishing.

The message of Mr. Onda’s presentation is with a lot of open data and tools available for everyone. Now is the time to create ample chances for the application of satellite data. It would be a challenge and beneficial to find ways to promote sustainable tourism using satellite data.

Japan’s Tourism Dynamics Observed through Data

In the second session of the event, it was presented by Asst. Prof. Kodaka Akira, Ph.D. (Graduate School of System Design and Management, Keio University) about “Japan’s Tourism Dynamics Observed through Data.” His presentation was opened with an explanation briefly on the condition of the tourist inbound-outbound in Japan during the COVID-19 pandemic. Because of the declining tourism activities, Japanese government put a program to incentivize more travels called “Go To Travel” campaign, which started on July 22nd, 2020, and this was suspended on December 28th, 2020, because of the increasing number of COVID-19 cases. But currently, the Japanese government is trying to resume this program soon, estimated in 2022. This research shows the five prefectures in Japan’s major tourist destinations, and what happens in these areas observed through satellite data of remote sensing and people mobility data during COVID-19 (Kodaka et al., 2022). Due to the pandemic situation and the campaign, if people want to travel to these five major prefectures that are prime destinations for tourism in Japan, the changes in the atmospheric condition; i.e., density of CO2, NO2, and PM2.5, can be observed through satellite data because of the changes of people movement and mobility as well as the pollution.

Asst. Prof. Kodaka also showed the example of the people’s movement through satellite data during an emergency decree in Japan. It shows that many people stayed in their homes/locations in Kyoto Prefecture during these times and abided by the emergency decree situation. However, after the emergency decree was lifted, there were some changes noticeable as people started moving around and people started to travel to Kyoto and other prefectures. Using the Google Earth Engine (GEE), this research analyzed and visualized the changes in an atmospheric condition caused by the movement of people during the COVID-19 pandemic and observed the levels of pollution changes across times in the major five prefectures. Furthermore, it analyzed if the people’s activities were controlled during the pandemic by showing if there were changes in economic activities as well during the changes in certain times. Analyzing the atmospheric condition based on the CO2 as well, it can show the mobility activities of certain transportation such as taxis, buses, and other public transportation or private owned those were crossing these prefectures during the Go To Travel campaign periods and show the increasing/decreasing number of activities of tourism in the area.

The future research possibility, the frustration, and the need to travel for tourists after COVID-19 ends, the government can put a program to incentivize these travels needs and what kind of activities are suitable for them taking into account sustainable tourism or green tourism based on the past observation of these satellite data.

Basic Twitter Data Retrieval for Analysis

The third session was presented about “Basic Twitter Data Retrieval for Analysis” by Jing Tang, D.Eng. (Lecturer of the Robotics and Artificial Intelligence Engineering Program, International School of Engineering, Faculty of Engineering; and Risk and Disaster Management Program, Graduate School, Chulalongkorn University; and a member of the Disaster and Risk Management Information Systems Research Unit, Chulalongkorn University). When people talk about Big Data as knowledge, it can separate these data mainly into two categories: machine-generated data and the second type is the data that is human generated. A perfect example of the machine-generated data is the satellite data that is explained in the previous presentation. Then, there is also human-generated data, such as the data on social media. This project mainly focused on this area. The Twitter application platform was selected because it generated a lot of data daily and is considered one of the most popular social media regarding humans’ generated data. In Thailand, Facebook took over the majority of users in social media, followed by Twitter daily. Compared with Facebook users, Twitter has increased the number of users in Thailand since 2017. One reason is that Facebook is used to share posts with related groups of people. On the other hand, people use Twitter mainly on the topics and trends mostly happening in real-time. Therefore, it would provide a more detailed and comprehensive analysis of specific issues; in this case, COVID-19 and Thailand tourism.

Twitter data-driven analysis can be separated into two types: the first one is variable-based analysis. It did not touch the tweet itself. Rather, it touches other variable aspects, such as the number of followers, likes, retweets, etc. Another type is content analysis, which focuses on the tweet itself. The analysis also can be separated furthermore into three types: The first one is Sentiment Analysis which focuses on the people opinion on a certain topic, whether they like or dislike the tweet or neutral; the second one is Topic Modelling, where it clusters the text into several groups and see what the correlation of the tweet to certain trends is and what the most popular topic recently is; the third type of the analysis is Identity Recognition, where it would find certain type information in the tweet regarding the identity such as location, name/username, time, activity name, country name, etc.

At the beginning period of the COVID-19 pandemic, there was research on the trends of tweets during the COVID-19 pandemic (Leelawat et al., 2020). This research was conducted during the early days of the pandemic, and it planned to explore if Twitter could identify what other names were used to describe the COVID-19. Due to the beginning phase, there was no official name at that moment. This research collected the data from January 2020-February 2020; it focuses on three languages, English, Japanese and Chinese tweets. The results show the alternative names of the COVID-19 in the tweets during these periods. Other topics related to the tweet were the news reported, such as the number of global cases, the number of deaths globally, other clustered instances, etc.

Another application of this research can be used on the tourism-related topics. For example, in Thailand’s case, it could be related to tweets on the Phuket Sandbox program or other related issues. These tweets can be collected and analyzed to make certain decisions for tourism benefits. Another example is related to sentiment analysis in Bangkok, Thailand, during the COVID-19 pandemic (Sontayasara et al., 2021). This study used the Support Vector Machine (SVM) algorithm to analyze the tweets. Furthermore, it can analyze the identity recognition to analyze what the tourist could be improved. Therefore, it can serve as a guideline to promote tourism to attract more people to Bangkok. We collected tweets related to Bangkok tourism during the study time. Then, the data was separated into positive, negative, or neutral tweets. The key finding of the research was that the positive tweets were related to food, city, temple, etc. Negative tweets were related to taxi drivers and flights. Neutral tweets were related to the temple, hotel, etc. Accordingly, the findings can be helpful for the government if they would like to promote and improve tourism policy related to the COVID-19 policy. They can use these specific keywords on what to improve.

Basic Twitter Data Retrieval for Analysis

The fourth session was presented by Mr. Kumpol Saengtabtim (Doctoral student of the Department of Industrial Engineering, Faculty of Engineering, Chulalongkorn University) on the topic of “Basic Twitter Data Retrieval for Analysis,” continued from Dr. Jing Tang’s. This session was started with the data analysis of the content from Twitter. The Python coding language connects it using the Twitter API and Tweepy. Twitter is a popular social media platform nowadays whose number of users is increasing continuously. The Twitter API can catch the data about current situation or tourism industry with the details of the retweet, favorite, time, and geolocation, from each post to conduct the analysis process. Also, keywords and hashtags can be emphasized in the trending search on the Twitter platform through the different interesting terms. Twitter API can retrieve data from specific users such as influencers or related organizations. The Twitter developer subscription is necessary for older data set. The Twitter developer subscription is necessary for older data set.

These data can be analyzed for the demand or opinion of the users/people to research for the weakness and strengths. He and DRMIS Research Unit also gained the data about the tourism destination in Thailand with the tourists’ opinion about the COVID-19 situation, which primarily involved their concerns about traveling to Thailand. In the implementation process, the planning is the first phase to specify the target study group and duration of the data collection. The second step is a coding process to collect the data as the plan. Then, we can move to the implementation phase for the details of the data collection. The next step is the analysis phase which is the objective of the research to find the opinion and the demand of the users.

Tourism Business Continuity Management

The last presentation in this seminar was presented by Asst. Prof. Natt Leelawat, D.Eng., MBCI (Assistant Professor of the Department of Industrial Engineering, Faculty of Engineering, Chulalongkorn University; Director of the Risk and Disaster Management Program, Graduate School, Chulalongkorn University; and the Head of the Disaster and Risk Management Information Systems Research Unit, Chulalongkorn University) in “Tourism Business Continuity Management.” The research results and methodology done in this project were introduced during the introduction. As a result, there are capable of implementing the tourism business risk management to maintain continuity. Furthermore, the risk matrix was introduced as the tool of the risk assessment.

His presentation focused on the business continuity management of the tourism business. Begin with the difference between BCM and BCP (Business Continuity Plan) which the BCP is the product of the BCM process through the risk assessment, business impact analysis, and other steps. BCM is to “make the business operate as usual on the acceptable level.” Thus, BCP is the instructions that need to operate even during the crisis and need to conduct the drills to find the gap of the development before back to another round of development again.

He introduced the advice from the international organizations, UNDRR and ADPC, which cooperated to develop the suggestions for the entrepreneur in BCM operation during a crisis. There are news and situation updates (government regulations are included), product and service understanding, communication with both staff and customers, social distancing regulations, staffs’ health protection, supply chain preparation, help request to the related organization, COVID-19 confronting protocol, etc.

Before the end of his presentation, Asst. Prof. Natt gave his opinion that the COVID-19 situation made the BCP details different from the other emergency due to the resources that affected the physical equipment to the human resources that need to be protected first.

Acknowledgement

This project is supported by the Special Program for Research Against COVID-19 (CU SPRAC 2101), JICA Project for AUN/SEED-Net. We would like to thank all DRMIS members and SDM members who helped and supported our project. Lastly, we would like to acknowledge a kind support from the Tourism Authority of Thailand.

References

Kodaka, A., Detera, B. J., Onda, Y., Leelawat, N., Tang, J., Laosunthara, A., Saengtabtim, K., & Kohtake, N. (2022). Interventions to support tourism and its impact on air quality – A case study of the Go To Travel campaign in Japan –. Journal of Disaster Research, 17 (1), 123-135. doi: 10.20965/jdr.2022.p0123

Kodaka, A., Leelawat, N., Tang, J., Onda, Y., Kohtake, N., Laosunthara, A., Saengtabtim, K., & Sochoeiya, P. (2021). Influential factors on aerosol change during COVID-19 in Ayutthaya, Thailand. Engineering Journal, 25 (8), 187-196. doi: 10.4186/ej.2021.25.8.187

Leelawat, N., Tang, J., Krutphong, K., Chaichanasiri, S., Kanno, T., Li, C. W., Le, L. T. Q., Dung, H. Q., Saengtabtim, K., & Laosunthara, A. (2021). Comparison of the initial overseas evacuation operations due to COVID-19: A focus on Asian countries. Journal of Disaster Research, 16 (7), 1137-1146. doi: 10.20965/jdr.2021.p1137

Leelawat, N., Tang, J., Saengtabtim, K., & Laosunthara, A. (2020). Trends of tweets on the Coronavirus Disease-2019 (COVID-19) pandemic. Journal of Disaster Research, 15 (4), 530-533. doi: 10.20965/jdr.2020.p0530

Sontayasara, T., Jariyapongpaiboon, S., Promjun, A., Seelpipat, N., Saengtabtim, K., Tang, J., & Leelawat, N. (2021). Twitter sentiment analysis on Bangkok tourism during the COVID-19 situation using support vector machine algorithm. Journal of Disaster Research, 16 (1), 24-30. doi: 10.20965/jdr.2021.p0024

For more information, please contact the Disaster and Risk Management Information Systems Research Unit, Chulalongkorn University.

URL: http://drmis.eng.chula.ac.th
Facebook: https://www.facebook.com/DRMISChula
Address: Room 511, 5F, Engineering Building 4, Department of Industrial Engineering, Faculty of Engineering, Chulalongkorn University, 254 Phayathai Road, Pathumwan, Bangkok 10330 Thailand.