近年來，社群平台的崛起更改了行銷產業的宣傳手法以及方式，而 Instagram 的興起，成為行銷的一大利器，品牌以及商家透過圖片以及貼文來推廣，已觸及更多的受眾。因此，熱門圖片與否將與行銷推廣有相互的影響。然而，於過往主要依靠經驗法則以及統計數據作為判斷，前者難以處理大量非結構化數據，後者易受到主觀影響。本研究旨在針對熱門美食圖片進行預測，有助於輔佐判斷行銷推廣後的成效預測。本研究主要對熱門美食圖片進行二元分類預測，研究範疇為新竹地區。於 API 爬蟲獲取資料後做資料擴增處理，並透過兩種 CNN 衍伸模型，分別為 EfficientNetB0 以及 ResNet50 來比較兩者模型於熱門美食圖片預測的表現。實驗結果顯示 EfficientNetB0 更具有優勢並表現優異的性能，以較少的訓練成本達到 96.28% 的準確度。此結果顯示本研究為熱門美食圖片分類預測提供有效的依據，於未來將可更進一步優化並且應用於實際的行銷領域。

In recent years, the rise of social media platforms has transformed marketing strategies and methods. In particular, Instagram has emerged as a powerful tool for marketing, as brands and businesses use images and posts to reach a wider audience. The popularity of images plays a significant role in marketing and promotion. However, in the past, decisions were mainly based on empirical rules and statistical data. Therefore, this study aims to predict the popularity of food images on
Instagram to assist in evaluating the effectiveness of marketing promotions. This research focuses on binary classification prediction for popular food images in the Hsinchu area. After collecting data through an API crawler, data augmentation is performed. EfficientNetB0 and ResNet50 are employed to compare their performance in predicting popular food images.
The experimental results demonstrate that EfficientNetB0 outperforms ResNet50 with excellent performance, achieving an accuracy of 96.28% with fewer training costs. This outcome indicates that this research provides an effective basis for predicting the classification of popular food images, which can be further optimized and applied in practical marketing applications in the future.

第一章 緒論 1
1.1 研究背景與動機 1
1.2 研究目的 4
1.3 研究架構 5
第二章 文獻回顧 7
2.1 社群媒體 7
2.1.1 社群媒體的發展與影響 7
2.1.2 網紅現象與市場行銷 7
2.2 網路爬蟲 8
2.2.1 爬蟲 8
2.2.2 反爬蟲 10
2.3 食物分類 11
2.4 熱門分類 12
2.5 資料擴增 14
2.6 模型架構 15
2.6.1 CNN 15
2.6.2 ResNet 19
2.6.3 EfficientNet 20
第三章 研究方法 22
3.1 研究流程 22
3.2 資料蒐集與爬蟲 23
3.3 資料擴增 24
3.4 模型架構 25
3.4.1 EfficientNetB0 25
3.4.2 ResNet50 27
3.5 模型評估 30
3.5.1 混淆矩陣 30
3.5.2 ROC 與 AUC 32
第四章 實驗結果與分析 34
4.1 實驗環境配置 34
4.2 問題闡述 35
4.3 資料集 35
4.4 資料擴增 36
4.5 實驗結果 36
4.5.1 模型訓練 37
4.5.2 模型訓練結果 38
4.5.3 模型訓練比較 44
第五章 結論與建議 47
5.1 結論 47
5.2 建議與未來發展 49
參考文獻 50

[1] Abousaleh, F.S., Cheng, W., Yu, N., & Tsao, Y. (2021). Multimodal Deep Learning Framework for Image Popularity Prediction on Social Media. IEEE Transactions on Cognitive and Developmental Systems, 13, 679-692.
[2] Abousaleh, F.S., Cheng, W., Yu, N., & Tsao, Y. (2021). Multimodal Deep Learning Framework for Image Popularity Prediction on Social Media. IEEE Transactions on Cognitive and Developmental Systems, 13, 679-692.
[3] adw0rd (2023) Instagrapi https://github.com/subzeroid/instagrapi.
[4] Astuti, B. (2018). Analysis on the Effect of Instagram Use on Consumer Purchase Intensity.
[5] Barbosa, L., & Freire, J. (2005). Searching for Hidden-Web Databases. International Workshop on the Web and Databases.
[6] Belanche, D., Flavián, M., & Ibáñez-Sánchez, S. (2020). Followers’ reactions to influencers’ Instagram posts. Spanish Journal of Marketing - ESIC, 24, 37-53.
[7] Bhatt, D., Patel, C.I., Talsania, H.N., Patel, J.A., Vaghela, R., Pandya, S., Modi, K.J., & Ghayvat, H. (2021). CNN Variants for Computer Vision: History, Architecture, Application, Challenges and Future Scope. Electronics.
[8] Bossard, L., Guillaumin, M., & Gool, L.V. (2014). Food-101 - Mining Discriminative Components with Random Forests. European Conference on Computer Vision.
[9] Bruhn, M., Schoenmueller, V. & Schäfer, D.B. (2012), Are social media replacing traditional media in terms of brand equity creation?, Management Research Review, Vol. 35 No. 9, 770-790.
[10] Casaló, L.V., Flavián, C., & Ibáñez-Sánchez, S. (2020). Influencers on Instagram: Antecedents and consequences of opinion leadership. Journal of Business Research, 117, 510-519.
[11] Chang, K.C., He, B., & Zhang, Z. (2005). Toward Large Scale Integration: Building a MetaQuerier over Databases on the Web. Conference on Innovative Data Systems Research.
[12] Che, J.W., Cheung, C.M., & Thadani, D.R. (2017). Consumer Purchase Decision in Instagram Stores: The Role of Consumer Trust. Hawaii International Conference on System Sciences.
[13] Ding, K., Ma, K., & Wang, S. (2019). Intrinsic Image Popularity Assessment. Proceedings of the 27th ACM International Conference on Multimedia.
[14] Djafarova, E., & Rushworth, C. (2017). Exploring the credibility of online celebrities' Instagram profiles in influencing the purchase decisions of young female users. Comput. Hum. Behav., 68, 1-7.
[15] Edwards, J., McCurley, K.S., & Tomlin, J.A. (2001). An adaptive model for optimizing performance of an incremental web crawler. The Web Conference.
[16] Gayberi, M., & Öğüdücü, Ş.G. (2019). Popularity Prediction of Posts in Social Networks Based on User, Post and Image Features. Proceedings of the 11th International Conference on Management of Digital EcoSystems.
[17] Haque, A., & Singh, S. (2015). Anti-scraping application development. 2015 International Conference on Advances in Computing, Communications and Informatics (ICACCI), 869-874.
[18] He, K., Zhang, X., Ren, S., & Sun, J. (2015). Deep Residual Learning for Image Recognition. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 770-778.
[19] Hridoy, R.H., Akter, F., Mahfuzullah, M.A., & Ferdowsy, F. (2021). A Computer Vision Based Food Recognition Approach for Controlling Inflammation to Enhance Quality of Life of Psoriasis Patients. 2021 International Conference on Information Technology (ICIT), 543-548.
https://www.statista.com/statistics/272014/global-social-networks-ranked-by-number-of-users/
[20] Jian-qiao, Y. (2012). Crawling Strategy of Deep Web Crawler. Computer Engineering.
[21] Joseph, N., Sultan, A., Kar, A.K., & Ilavarasan, P.V. (2018). Machine Learning Approach to Analyze and Predict the Popularity of Tweets with Images. IFIP International Conference on e-Business, e-Services, and e-Society.
[22] Kim, S., Han, J., Yoo, S., & Gerla, M. (2017). How Are Social Influencers Connected in Instagram?. In: Ciampaglia, G., Mashhadi, A., Yasseri, T. (eds) Social Informatics. SocInfo 2017. Lecture Notes in Computer Science, vol 10540. Springer, Cham.
[23] Krizhevsky, A., Sutskever, I., & Hinton, G.E. (2012). ImageNet classification with deep convolutional neural networks. Communications of the ACM, 60, 84 - 90.
[24] Kumar M. & Shanker S.(2022).An Analysis of Instagram Algorithms and New updates in 2022. VOL. 13 NO. SPECIAL ISSUE (2022): TRINITY INSTITUTE OF PROFESSIONAL STUDIES (TIPSCON 2022)
[25] LeCun, Y., Boser, B.E., Denker, J.S., Henderson, D., Howard, R.E., Hubbard, W.E., & Jackel, L.D. (1989). Backpropagation Applied to Handwritten Zip Code Recognition. Neural Computation, 1, 541-551.
[26] Lim, X.J., Radzol, A.R., Cheah, J., & Wong, M.W. (2017). The Impact of Social Media Influencers on Purchase Intention and the Mediation Effect of Customer Attitude.
[27] Mangold, W.G., & Faulds, D.J. (2009). Social media: The new hybrid element of the promotion mix. Business Horizons, 52, 357-365.
[28] Mikołajczyk, A., & Grochowski, M. (2018). Data augmentation for improving deep learning in image classification problem. 2018 International Interdisciplinary PhD Workshop (IIPhDW), 117-122.
[29] Rajayogi, J.R., Manjunath, G.R., & Shobha, G. (2019). Indian Food Image Classification with Transfer Learning. 2019 4th International Conference on Computational Systems and Information Technology for Sustainable Solution (CSITSS), 4, 1-4.
[30] Sharma, S., & Gupta, P.J. (2015). The anatomy of web crawlers. International Conference on Computing, Communication & Automation, 849-853.
[31] Shestakov, D. (2011). Databases on the web: national web domain survey. International Database Engineering and Applications Symposium.
[32] Shupei Yuan & Chen Lou (2020) How Social Media Influencers Foster Relationships with Followers: The Roles of Source Credibility and Fairness in Parasocial Relationship and Product Interest, Journal of Interactive Advertising, 20:2, 133-147.
[33] Sokolova, K., & Kefi, H. (2020). Instagram and YouTube bloggers promote it, why should I buy? How credibility and parasocial interaction influence purchase intentions. Journal of Retailing and Consumer Services, 53.
[34] Sriram K V, Namitha KP, & Giridhar B Kamath. (2021). Social media advertisements and their influence on consumer purchase intention. Cogent Business & Management, 8.
[35] statista. (2023). Most popular social networks worldwide as of January 2023, ranked by number of monthly active users
[36] Sultana, J., Ahmed, B.M., Masud, M.M., Huq, A.K., Ali, M.E., & Naznin, M. (2023). A Study on Food Value Estimation From Images: Taxonomies, Datasets, and Techniques. IEEE Access, 11, 45910-45935.
[37] Tan, M., & Le, Q.V. (2019). EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks. ArXiv, abs/1905.11946.
[38] Tateno, M., Teo, A.R., Ukai, W., Kanazawa, J., Katsuki, R., Kubo, H., & Kato, T.A. (2019). Internet Addiction, Smartphone Addiction, and Hikikomori Trait in Japanese Young Adult: Social Isolation and Social Network. Frontiers in Psychiatry, 10.
[39] teresawang(2022).不同的爬蟲種類。IT邦幫忙。https://ithelp.ithome.com.tw/articles/10297404
[40] Wang, A., & Gao, X. (2019). Multifunctional Product Marketing Using Social Media Based on the Variable-Scale Clustering. Tehnicki vjesnik - Technical Gazette.
[41] Zahisham, Z., Lee, C.P., & Lim, K.M. (2020). Food Recognition with ResNet-50. 2020 IEEE 2nd International Conference on Artificial Intelligence in Engineering and Technology (IICAIET), 1-5.
[42] Zhao, F., Zhou, J., Nie, C., Huang, H., & Jin, H. (2016). SmartCrawler: A Two-Stage Crawler for Efficiently Harvesting Deep-Web Interfaces. IEEE Transactions on Services Computing, 9, 608-620.