系統識別號 | U0002-2701202112151400 |
---|---|
DOI | 10.6846/TKU.2021.00730 |
論文名稱(中文) | 使用機器學習建立基於圖像資料的推薦系統─以新加坡旅遊為例 |
論文名稱(英文) | Photos-based recommendation system with the use of Machine Learning: A case study of Singapore tourism |
第三語言論文名稱 | |
校院名稱 | 淡江大學 |
系所名稱(中文) | 經營管理全英語碩士學位學程 |
系所名稱(英文) | Master's Program in Business and Management (English-Taught Program) |
外國學位學校名稱 | |
外國學位學院名稱 | |
外國學位研究所名稱 | |
學年度 | 109 |
學期 | 1 |
出版年 | 110 |
研究生(中文) | 呂明明 |
研究生(英文) | Chanyanat Soontornsittirat |
學號 | 607585402 |
學位類別 | 碩士 |
語言別 | 英文 |
第二語言別 | |
口試日期 | 2021-01-12 |
論文頁數 | 38頁 |
口試委員 |
指導教授
-
吳家齊(arthurwu@mail.tku.edu.tw)
委員 - 廖建翔(052122@mail.fju.edu.tw) 委員 - 方郁惠(yhfang@mail.tku.edu.tw) |
關鍵字(中) |
推薦系統 機器學習 照片分析 新加坡旅遊 |
關鍵字(英) |
Recommendation System Machine Learning Photo Analysis Singapore Tourism |
第三語言關鍵字 | |
學科別分類 | |
中文摘要 |
在過去十年中,全球休閒旅遊和商務旅遊的市場逐漸增加,並且預計 會在下一個十年持續增長。現有研究指出,有超過一半的旅行者會在 旅行前上網收集資訊,並且根據網路資源規劃行程。然而,由於資訊 過多,反而讓使用者很難篩選符合偏好的資訊並決定遊覽景點。 本研究聚焦於,如何僅基於Instagram用戶所上傳的照片,構建基於照 片的有效推薦系統。關於景點(Point of Interests, POIs)的特徵,我 們從Visit Singapore和TripAdvisor兩個網站所收集的資料中萃取。我們 所提出的方法分為兩個階段,第一階段嘗試通過匹配使用者和景點的 照片特徵標籤來決定要向使用者推薦的景點類別。在第二階段中,我 們根據每個景點的整體評分、評論數量,和熱門關鍵字與使用者的匹 配決定每個景點對於個別使用者的分數。並在依據分數排序後,從第 一階段所挑選出的類別中推薦排名前面的景點。 實驗結果顯示,即使使用先進的人工智慧工具進行照片和圖形分析, 其有效性與準確性仍然不如人類自行提供的資訊。然而,在無法取得 使用者資訊或資訊不足時,照片和圖形分析仍可帶來可接受的效果。 |
英文摘要 |
Worldwide leisure tourism and business tourism have been gradually increased over the past ten years and aim to continue growing for the next decade. Previous research shows that more than half of travelers would do pre-travel research and decide where to go based on the available source online. However, due to the information overload, it is difficult for them to obtain the right information that suits their preferences and find the place to visit. This study aims to demonstrate that a photos-based recommendation system can be effectively built based solely on users’ Instagram photos. For Point of Interests (POIs) data, we use information from Visit Singapore and TripAdvisor. Our method divides into two major phases, first phase, is where we try to determine which categories to recommend to users by matching Google Vision API’ labels from users and POIs. In our second phase, we give each POIs score from the rating, the number of reviews, and popular mentions match with users’ labels. We then rank POIs and recommend the top POIs from their phase one categories. The result shows that even with a well-developed artificial intelligence for photo and graph analysis, it still cannot perform as effectively and as accurately as human-provided information. However, photo and graph analysis are still very useful in various situation when the information is insufficient. |
第三語言摘要 | |
論文目次 |
Acknowledgement...................................I 論文提要內容.......................................II Abstract..........................................III Table of Contents.................................V List of Tables................................... VI List of Figures...................................VII 1. Introduction...................................1 2. Literature Review .............................7 2.1 Users’ Profiling..............................7 2.2 Recommendation System.........................8 2.2.1 Content-based filtering.....................8 2.2.2 Collaborative filtering.....................9 2.2.3 Hybrid filtering ...........................9 2.2.4 Tourism recommendation system...............10 3. Methodology....................................11 3.1 Data collection and pre-processing............13 3.1.1 Point of interest in Singapore..............13 3.1.2 Users.......................................15 3.2 Proposed Method...............................15 3.2.1 Phase one...................................16 3.2.2 Phase two ..................................19 4. Experimental ..................................22 4.1 Experiment Design.............................22 4.2 Evaluation result.............................25 4.3 Real Case.....................................27 5. Conclusion and Suggestions.....................31 References........................................33 List of Tables Table 3.1 POIs Category ..........................14 Table 3.2 Labels Example .........................18 Table 4.1 Sub-categories frequency.........................................24 Table 4.2 Popular Mentions Match..................24 Table 4.3 Experimental results....................26 Table 4.4 User 10 Photo Labels Example............28 Table 4.5 Phase two real case: User 10 ...........29 Table 4.6 User 10 Final Recommendation............30 List of Figures Figure 1.1 Leisure and business tourism spending worldwide from .......1 Figure 1.2 Total number of International Visitors came to Singapore....3 Figure 3.1 Methodology Diagram...................12 Figure 3.2 Phase one Methodology Diagram.........16 |
參考文獻 |
Aslam, S. (2020). Instagram by the Numbers: Stats, Demographics & Fun Facts. Retrieved October 13, 2020, from https://www.omnicoreagency.com/instagram-statistics/ Blanco-Fernandez, Y., Pazos-Arias, J. J., Gil-Solla, A., Ramos-Cabrer, M., & Lopez-Nores, M. (2008). Providing entertainment by content-based filtering and semantic reasoning in intelligent recommender systems. IEEE Transactions on Consumer Electronics, 54(2), 727-735. Chen, W., Niu, Z., Zhao, X., & Li, Y. (2014). A hybrid recommendation algorithm adapted in e-learning environments. World Wide Web, 17(2), 271- 284. Clement, J. (2020, October 29). Instagram: Age and gender demographics. Retrieved January 16, 2021, from https://www.statista.com/statistics/248769/age-distribution-of-worldwide- instagram-users/ Ferwerda, B., & Tkalcic, M. (2018, July). Predicting users' personality from instagram pictures: Using visual and/or content features?. In Proceedings of the 26th Conference on User Modeling, Adaptation and Personalization (pp. 157-161). Ghauth, K. I., & Abdullah, N. A. (2010). Learning materials recommendation using good learners’ ratings and content-based filtering. Educational technology research and development, 58(6), 711-727. Intelligence Unit, The Economist (2019). Business Environment Rankings Which country is best to do business in? Retrieved November 09, 2020, from http://www.iberglobal.com/files/business_climate_eiu.pdf Jia, Z., Yang, Y., Gao, W., & Chen, X. (2015, February). User-based collaborative filtering for tourist attraction recommendations. In 2015 IEEE International Conference on Computational Intelligence & Communication Technology (pp. 22-25). IEEE. Kabassi, K. (2010). Personalizing recommendations for tourists. Telematics and Informatics, 27(1), 51-66. Konstan, J. A., Miller, B. N., Maltz, D., Herlocker, J. L., Gordon, L. R., & Riedl, J. (1997). GroupLens: applying collaborative filtering to Usenet news. Communications of the ACM, 40(3), 77-87. Koren, Y., & Bell, R. (2015). Advances in collaborative filtering. In Recommender systems handbook (pp. 77-118). Springer, Boston, MA. Lekakos, G., & Caravelas, P. (2008). A hybrid approach for movie recommendation. Multimedia tools and applications, 36(1-2), 55-70. Li, G., Zhu, T., Hua, J., Yuan, T., Niu, Z., Li, T., & Zhang, H. (2019). Asking images: Hybrid recommendation system for tourist spots by hierarchical sampling statistics and multimodal visual Bayesian personalized ranking. IEEE Access, 7, 126539-126560. Linden, G., Smith, B., & York, J. (2003). Amazon. com recommendations: Item-to-item collaborative filtering. IEEE Internet computing, 7(1), 76-80 Liu, L., Preotiuc-Pietro, D., Samani, Z. R., Moghaddam, M. E., & Ungar, L. H. (2016, March). Analyzing personality through social media profile picture choice. In ICWSM (pp. 211-220). Lock, P. (2020, February 10). Business travel spending worldwide 2000- 2019. Retrieved October 13, 2020, from https://www.statista.com/statistics/1093295/business-travel-spending- worldwide/ Lock, P. (2020, February 10). Leisure travel spending worldwide 2000- 2019. Retrieved October 13, 2020, from https://www.statista.com/statistics/1093335/leisure-travel-spending- worldwide/ Logesh, R., & Subramaniyaswamy, V. (2019). Exploring hybrid recommender systems for personalized travel applications. In Cognitive informatics and soft computing (pp. 535-544). Springer, Singapore. Lops, P., De Gemmis, M., & Semeraro, G. (2011). Content-based recommender systems: State of the art and trends. In Recommender systems handbook (pp. 73-105). Springer, Boston, MA. Nie, J., Huang, L., Cui, P., Li, Z., Yan, Y., Wei, Z., & Zhu, W. (2016, September). Social media profiler: Inferring your social media personality from visual attributes in portrait. In Pacific Rim Conference on Multimedia (pp. 640-649). Springer, Cham. Miller, P. (2020, May 11). The World's Best Airports in 2020 are announced. Retrieved October 13, 2020, from https://www.worldairportawards.com/the-worlds-best-airports-in-2020-are- announced/ Pazzani, M. J., & Billsus, D. (2007). Content-based recommendation systems. In The adaptive web (pp. 325-341). Springer, Berlin, Heidelberg. Renjith, S., & Anjali, C. (2013). A personalized travel recommender model based on content-based prediction and collaborative recommendation. International Journal of Computer Science and Mobile Computing, ICMIC13, 66-73. Robino, D. (2019). Global Destination Cities Index 2019. Retrieved October 13, 2020, from https://newsroom.mastercard.com/wp- content/uploads/2019/09/GDCI-Global-Report-FINAL1.pdf Schein, A. I., Popescul, A., Ungar, L. H., & Pennock, D. M. (2002, August). Methods and metrics for cold-start recommendations. In Proceedings of the 25th annual international ACM SIGIR conference on Research and development in information retrieval (pp. 253-260). Sertkan, M., Neidhardt, J., & Werthner, H. (2020). From Pictures to Travel Characteristics: Deep Learning-Based Profiling of Tourists and Tourism Destinations. In Information and Communication Technologies in Tourism 2020 (pp. 142-153). Springer, Cham. Singapore Department Of Statistics: SingStat Table Builder - Variables/Time Period Selection. (2019). Retrieved October 13, 2020, from https://www.tablebuilder.singstat.gov.sg/publicfacing/createDataTable.actio n?refId=1531 Singapore is second-safest country for travellers in 2020 ranking. (2020, February 21). Retrieved October 13, 2020, from https://www.thepeakmagazine.com.sg/gourmet-travel/singapore-second- safest-country-for-travellers-2020-ranking/ Son, J., & Kim, S. B. (2017). Content-based filtering for recommendation systems using multiattribute networks. Expert Systems with Applications, 89, 404-412. The Four Booking Behaviours Driving Travellers. (2016, October 17). Retrieved October 13, 2020, from https://www.tripadvisor.co.uk/TripAdvisorInsights/w797?m=21588 Traffic Statistics: Changi Airport Group. (2019). Retrieved October 13, 2020, from https://www.changiairport.com/corporate/our-expertise/air- hub/traffic-statistics.html Who Will Finance Innovation? (n.d.). Retrieved November 08, 2020, from https://www.globalinnovationindex.org/home Yeh, D. Y., & Cheng, C. H. (2015). Recommendation system for popular tourist attractions in Taiwan using Delphi panel and repertory grid techniques. Tourism Management, 46, 164-176. Chicago |
論文全文使用權限 |
如有問題,歡迎洽詢!
圖書館數位資訊組 (02)2621-5656 轉 2487 或 來信