1. NCCU Academic Hub
  2. Publications
  3. College of Informatics
  4. Theses
  5. Item

Publications-Theses

Article View/Open

Publication Export

Google ScholarTM

NCCU Library

Citation Infomation

Related Publications in TAIR

題名 數位行銷中無Cookie內容對於廣告投遞的產業推薦
Industry Recommendation for Cookieless Contents in Digital Marketing
作者 邱盈儒
Chiu, Ying-Ju
貢獻者 李蔡彥
Li, Tsai-Yen
邱盈儒
Chiu, Ying-Ju
關鍵詞 Cookieless
數位廣告
推薦系統
點擊率預測
自然語言處理
關鍵字萃取
Cookieless
Digital Advertising
Recommendation System
CTR Prediction
Natural Language Processing
Keyword Extraction
日期 2024
上傳時間 4-Sep-2024 14:35:54 (UTC+8)
摘要 隨著各個主流瀏覽器相繼宣布逐步淘汰第三方Cookie，數位廣告投放面臨巨大的挑戰。在遵守全球各國隱私法規的同時，如何仍能保持高精準度的廣告投放，成為數位廣告商的當務之急。本研究提出了一種在不使用任何使用者個人資料的情況下，通過解析使用者瀏覽的文本內容來進行廣告投放的潛在替代方法。我們採用了先進的深度學習技術來處理和理解文本內容。通過對使用者瀏覽的文章、新聞和其他文本內容進行語義分析，我們可以推測出使用者的潛在興趣和需求，進而實現精準的廣告推薦。 為了驗證此方法的有效性，我們進行了一系列實驗，重點測試了該方法在點擊率預測任務中的表現。實驗結果顯示，儘管不使用傳統的使用者行為數據，我們的方法仍能達到令人滿意的預測精度。這表明，通過解析文本內容，可以在一定程度上替代Cookie所提供的功能，為數位廣告商提供了一種可行的解決方案。此外，我們還比較了不同深度學習模型和參數配置對預測效果的影響，找出了在不同情境下的最佳配置。 本研究不僅為數位廣告投遞提供了一種新的思路，還展示了在無Cookie環境下利用文本內容進行廣告推薦的潛力。隨著數位行銷生態系統的不斷變化，我們的方法有望成為廣告商適應新形勢的重要工具，既能滿足隱私保護的需求，又能保持廣告的高效投放。
As major web browsers gradually announce the phasing out of third-party cookies, digital advertising faces significant challenges. Ensuring high-precision advertisement targeting while complying with global privacy regulations has become a critical issue for digital advertisers. This study proposes a potential alternative method for advertisement delivery without using any personal user data by analyzing the text content browsed by users. We employ advanced deep learning techniques to process and understand the text content. By semantically analyzing articles, news, and other textual content browsed by users, we can infer their potential interests and needs, thereby achieving precise advertisement recommendations. To validate the effectiveness of this method, we conducted a series of experiments, focusing on its performance in the click-through rate (CTR) prediction task. The experimental results show that our method can achieve satisfactory prediction accuracy even without traditional user behavior data. This indicates that text content analysis can partially replace the functionality provided by cookies, offering a feasible solution for digital advertisers. Additionally, we compared the effects of different deep learning models and parameter configurations on prediction performance, identifying the optimal setups under various scenarios. This study not only provides a new approach for digital advertisement delivery but also demonstrates the potential of using text content for advertisement recommendations in a cookieless environment. As the digital marketing ecosystem continues to evolve, our method is expected to become an essential tool for advertisers to adapt to new conditions, meeting privacy protection requirements while maintaining efficient advertisement delivery.
參考文獻 [1] A. Bahirat. Contextual recommendations using nlp in digital marketing. In X.-S. Yang, S. Sherratt, N. Dey, and A. Joshi, editors, Proceedings of Sixth International Congress on Information and Communication Technology, pages 655–664, Singa- pore, 2022. Springer Singapore. [2] S.Bharati,M.R.H.Mondal,P.Podder,andV.S.Prasath.Federatedlearning:Appli- cations, challenges and future directions. International Journal of Hybrid Intelligent Systems, 18(1–2):19–35, May 2022. [3] J. Carbonell and J. Goldstein. The use of mmr, diversity-based reranking for re- ordering documents and producing summaries. In Proceedings of the 21st Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR ’98, page 335–336, New York, NY, USA, 1998. Association for Computing Machinery. [4] H.-T. Cheng, L. Koc, J. Harmsen, T. Shaked, T. Chandra, H. Aradhye, G. Anderson, G. Corrado, W. Chai, M. Ispir, R. Anil, Z. Haque, L. Hong, V. Jain, X. Liu, and H. Shah. Wide & deep learning for recommender systems, 2016. [5] cropgpt. Confusion matrix – explanation, 2020. https://cropgpt.ai/ confusion-matrix-explanation/. [6] J. Devlin, M.-W. Chang, K. Lee, and K. Toutanova. Bert: Pre-training of deep bidi- rectional transformers for language understanding, 2019. [7] O. Espejel. how-to-train-sentence-transformers, 2022. https://huggingface. co/blog/how-to-train-sentence-transformers. [8] Google. Building a more private web: A path towards making third party cookies obsolete, 2020. https://blog.chromium.org/2020/01/ building-more-private-web-path-towards.html. [9] Google. The next step toward phasing out third-party cook- ies in chrome, 2023. https://blog.google/products/chrome/ privacy-sandbox-tracking-protection/. [10] Google. A new path for privacy sandbox on the web, 2024. https:// privacysandbox.com/news/privacy-sandbox-update/. [11] M. Grootendorst. Keybert: Minimal keyword extraction with bert., 2020. [12] H. Guo, R. Tang, Y. Ye, Z. Li, and X. He. Deepfm: A factorization-machine based neural network for ctr prediction, 2017. [13] S. Ioffe and C. Szegedy. Batch normalization: Accelerating deep network training by reducing internal covariate shift, 2015. [14] A. Jain, M. Pathak, and M. Divya Prabha. Tackling cookieless domain recommen- dation for digital advertising targetting. page 111–112, 2022. [15] S.khan,Q.M.Ilyas,andW.Anwar.Contextualadvertisingusingkeywordextraction through collocation. In Proceedings of the 7th International Conference on Frontiers of Information Technology, FIT ’09, New York, NY, USA, 2009. Association for Computing Machinery. [16] D. P. Kingma and J. Ba. Adam: A method for stochastic optimization, 2017. [17] J. Lian, X. Zhou, F. Zhang, Z. Chen, X. Xie, and G. Sun. xdeepfm: Combining explicit and implicit feature interactions for recommender systems. In Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, KDD '18. ACM, July 2018. [18] S. Lloyd. Least squares quantization in pcm. IEEE Transactions on Information Theory, 28(2):129–137, 1982. [19] L. Long. Practice papers effective first-party data collection in a privacy-first world. Applied Marketing Analytics, 7(3):202–210, 2022. [20] MartinThoma. Receiver operating characteristic (roc) curve, 2018. https:// commons.wikimedia.org/w/index.php?curid=70212136. [21] T. Mikolov, K. Chen, G. Corrado, and J. Dean. Efficient estimation of word repre- sentations in vector space, 2013. [22] M. Pachilakis, P. Papadopoulos, E. P. Markatos, and N. Kourtellis. No more chasing waterfalls: A measurement study of the header bidding ad-ecosystem. In Proceedings of the Internet Measurement Conference, IMC ’19, page 280–293, New York, NY, USA, 2019. Association for Computing Machinery. [23] Rakesh4realg. Neighborhood based collaborative filter- ing —part 4, 2019. https://medium.com/fnplus/ neighbourhood-based-collaborative-filtering-4b7caedd2d11. [24] N. Reimers and I. Gurevych. Sentence-bert: Sentence embeddings using siamese bert-networks, 2019. [25] S. Rendle. Factorization machines. In 2010 IEEE International Conference on Data Mining, pages 995–1000, 2010. [26] W. Shen. Deepctr: Easy-to-use,modular and extendible package of deep-learning based ctr models. https://github.com/shenweichen/deepctr, 2017. [27] W.Song,C.Shi,Z.Xiao,Z.Duan,Y.Xu,M.Zhang,andJ.Tang.Autoint:Automatic feature interaction learning via self-attentive neural networks. In Proceedings of the 28th ACM International Conference on Information and Knowledge Management, CIKM '19. ACM, Nov. 2019. [28] R. van Meteren. Using content-based filtering for recommendation. 2000. [29] A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. Kaiser, and I. Polosukhin. Attention is all you need, 2023. [30] J.Wang,W.Zhang,andS.Yuan.Displayadvertisingwithreal-timebidding(rtb)and behavioural targeting. Foundations and Trends in Information Retrieval, 11(4-5):297 –435, 2017. Cited by: 75; All Open Access, Green Open Access. [31] L. Wang, K.-C. Lee, and Q. Lu. Improving advertisement recommendation by en- riching user browser cookie attributes. volume 24-28-October-2016, page 2401 – 2404, 2016. [32] R.Wang,B.Fu,G.Fu,andM.Wang.Deep&crossnetworkforadclickpredictions, 2017. [33] R. Wang, R. Shivanna, D. Cheng, S. Jain, D. Lin, L. Hong, and E. Chi. Dcn v2: Improved deep & cross network and practical lessons for web-scale learning to rank systems. In Proceedings of the Web Conference 2021, WWW '21. ACM, Apr. 2021. [34] Wikipedia.Cosinesimilarity,2007.https://en.wikipedia.org/wiki/Cosine_ similarity. [35] Wikipedia. Kullback–leibler divergence, 2011. https://en.wikipedia.org/ wiki/Kullback%E2%80%93Leibler_divergence. [36] Wikipedia. Elbow method (clustering), 2016. https://en.wikipedia.org/ wiki/Elbow_method_(clustering). [37] Wikipedia. General data protection regulation, 2016. https://en.wikipedia. org/wiki/General_Data_Protection_Regulation. [38] Wikipedia. California consumer privacy act, 2018. https://en.wikipedia.org/ wiki/California_Consumer_Privacy_Act. [39] R. Zhang, Q.-d. Liu, Chun-Gui, J.-X. Wei, and Huiyi-Ma. Collaborative filtering for recommender systems. In 2014 Second International Conference on Advanced Cloud and Big Data, pages 301–308, 2014. [40] Y. Zhang. An introduction to matrix factorization and factorization machines in recommendation system, and beyond, 2022.
描述 碩士
國立政治大學
資訊科學系碩士在職專班
110971015
資料來源 http://thesis.lib.nccu.edu.tw/record/#G0110971015
資料類型 thesis
dc.contributor.advisor 李蔡彥zh_TW
dc.contributor.advisor Li, Tsai-Yenen_US
dc.contributor.author (Authors) 邱盈儒zh_TW
dc.contributor.author (Authors) Chiu, Ying-Juen_US
dc.creator (作者) 邱盈儒zh_TW
dc.creator (作者) Chiu, Ying-Juen_US
dc.date (日期) 2024en_US
dc.date.accessioned 4-Sep-2024 14:35:54 (UTC+8)-
dc.date.available 4-Sep-2024 14:35:54 (UTC+8)-
dc.date.issued (上傳時間) 4-Sep-2024 14:35:54 (UTC+8)-
dc.identifier (Other Identifiers) G0110971015en_US
dc.identifier.uri (URI) https://nccur.lib.nccu.edu.tw/handle/140.119/153281-
dc.description (描述) 碩士zh_TW
dc.description (描述) 國立政治大學zh_TW
dc.description (描述) 資訊科學系碩士在職專班zh_TW
dc.description (描述) 110971015zh_TW
dc.description.abstract (摘要) 隨著各個主流瀏覽器相繼宣布逐步淘汰第三方Cookie，數位廣告投放面臨巨大的挑戰。在遵守全球各國隱私法規的同時，如何仍能保持高精準度的廣告投放，成為數位廣告商的當務之急。本研究提出了一種在不使用任何使用者個人資料的情況下，通過解析使用者瀏覽的文本內容來進行廣告投放的潛在替代方法。我們採用了先進的深度學習技術來處理和理解文本內容。通過對使用者瀏覽的文章、新聞和其他文本內容進行語義分析，我們可以推測出使用者的潛在興趣和需求，進而實現精準的廣告推薦。 為了驗證此方法的有效性，我們進行了一系列實驗，重點測試了該方法在點擊率預測任務中的表現。實驗結果顯示，儘管不使用傳統的使用者行為數據，我們的方法仍能達到令人滿意的預測精度。這表明，通過解析文本內容，可以在一定程度上替代Cookie所提供的功能，為數位廣告商提供了一種可行的解決方案。此外，我們還比較了不同深度學習模型和參數配置對預測效果的影響，找出了在不同情境下的最佳配置。 本研究不僅為數位廣告投遞提供了一種新的思路，還展示了在無Cookie環境下利用文本內容進行廣告推薦的潛力。隨著數位行銷生態系統的不斷變化，我們的方法有望成為廣告商適應新形勢的重要工具，既能滿足隱私保護的需求，又能保持廣告的高效投放。zh_TW
dc.description.abstract (摘要) As major web browsers gradually announce the phasing out of third-party cookies, digital advertising faces significant challenges. Ensuring high-precision advertisement targeting while complying with global privacy regulations has become a critical issue for digital advertisers. This study proposes a potential alternative method for advertisement delivery without using any personal user data by analyzing the text content browsed by users. We employ advanced deep learning techniques to process and understand the text content. By semantically analyzing articles, news, and other textual content browsed by users, we can infer their potential interests and needs, thereby achieving precise advertisement recommendations. To validate the effectiveness of this method, we conducted a series of experiments, focusing on its performance in the click-through rate (CTR) prediction task. The experimental results show that our method can achieve satisfactory prediction accuracy even without traditional user behavior data. This indicates that text content analysis can partially replace the functionality provided by cookies, offering a feasible solution for digital advertisers. Additionally, we compared the effects of different deep learning models and parameter configurations on prediction performance, identifying the optimal setups under various scenarios. This study not only provides a new approach for digital advertisement delivery but also demonstrates the potential of using text content for advertisement recommendations in a cookieless environment. As the digital marketing ecosystem continues to evolve, our method is expected to become an essential tool for advertisers to adapt to new conditions, meeting privacy protection requirements while maintaining efficient advertisement delivery.en_US
dc.description.tableofcontents 第一章 緒論 1 1.1 研究背景與動機 1 1.1.1 Cookie介紹 1 1.1.1.1 第一方Cookie（1st Party Cookie） 1 1.1.1.2 第三方Cookie（3rd Party Cookie） 2 1.1.2 研究動機 2 1.2 研究目的 3 1.3 論文架構 4 第二章 相關研究與技術背景 6 2.1 數位行銷生態系統 6 2.2 數位廣告投放 7 2.2.1 數位廣告投放之策略 8 2.2.2 相關研究 9 2.3 數位行銷指標 9 2.4 數位行銷中常見的演算法 11 2.4.1 數位行銷與機器學習 11 2.4.2 數位行銷與深度學習 13 2.4.2.1 深度學習模型的優勢 19 2.4.2.2 實際應用案例 20 2.5 評估指標 21 2.5.1 混淆矩陣 21 2.5.2 ROC與AUC 22 2.5.3 LogLoss 23 第三章 研究方法 25 3.1 基本構想 25 3.2 前期研究 25 3.2.1 資料集 26 3.2.2 資料前處理 26 3.2.3 深度學習推薦模型 27 3.3 研究架構設計 28 3.3.1 問題陳述 28 3.3.2 研究架構 29 3.4 目標設定 29 第四章 研究過程與實驗結果分析 32 4.1 實驗環境 32 4.2 研究過程 32 4.2.1 資料集前處理 32 4.2.2 文本特徵工程 33 4.2.3 資料集實驗與觀察 34 4.2.4 模型實驗 37 4.2.4.1 實驗一：固定超參數 37 4.2.4.2 實驗二：GridSearchCV找尋最佳超參數配置 38 4.2.4.3 實驗三：測試不同資料區間 40 4.2.4.4 實驗四：模型推論 43 4.3 研究結果分析 45 第五章 結論與未來研究方向 51 5.1 結論 51 5.2 未來研究方向 52 參考文獻 56zh_TW
dc.format.extent 3091497 bytes-
dc.format.mimetype application/pdf-
dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0110971015en_US
dc.subject (關鍵詞) Cookielesszh_TW
dc.subject (關鍵詞) 數位廣告zh_TW
dc.subject (關鍵詞) 推薦系統zh_TW
dc.subject (關鍵詞) 點擊率預測zh_TW
dc.subject (關鍵詞) 自然語言處理zh_TW
dc.subject (關鍵詞) 關鍵字萃取zh_TW
dc.subject (關鍵詞) Cookielessen_US
dc.subject (關鍵詞) Digital Advertisingen_US
dc.subject (關鍵詞) Recommendation Systemen_US
dc.subject (關鍵詞) CTR Predictionen_US
dc.subject (關鍵詞) Natural Language Processingen_US
dc.subject (關鍵詞) Keyword Extractionen_US
dc.title (題名) 數位行銷中無Cookie內容對於廣告投遞的產業推薦zh_TW
dc.title (題名) Industry Recommendation for Cookieless Contents in Digital Marketingen_US
dc.type (資料類型) thesisen_US
dc.relation.reference (參考文獻) [1] A. Bahirat. Contextual recommendations using nlp in digital marketing. In X.-S. Yang, S. Sherratt, N. Dey, and A. Joshi, editors, Proceedings of Sixth International Congress on Information and Communication Technology, pages 655–664, Singa- pore, 2022. Springer Singapore. [2] S.Bharati,M.R.H.Mondal,P.Podder,andV.S.Prasath.Federatedlearning:Appli- cations, challenges and future directions. International Journal of Hybrid Intelligent Systems, 18(1–2):19–35, May 2022. [3] J. Carbonell and J. Goldstein. The use of mmr, diversity-based reranking for re- ordering documents and producing summaries. In Proceedings of the 21st Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR ’98, page 335–336, New York, NY, USA, 1998. Association for Computing Machinery. [4] H.-T. Cheng, L. Koc, J. Harmsen, T. Shaked, T. Chandra, H. Aradhye, G. Anderson, G. Corrado, W. Chai, M. Ispir, R. Anil, Z. Haque, L. Hong, V. Jain, X. Liu, and H. Shah. Wide & deep learning for recommender systems, 2016. [5] cropgpt. Confusion matrix – explanation, 2020. https://cropgpt.ai/ confusion-matrix-explanation/. [6] J. Devlin, M.-W. Chang, K. Lee, and K. Toutanova. Bert: Pre-training of deep bidi- rectional transformers for language understanding, 2019. [7] O. Espejel. how-to-train-sentence-transformers, 2022. https://huggingface. co/blog/how-to-train-sentence-transformers. [8] Google. Building a more private web: A path towards making third party cookies obsolete, 2020. https://blog.chromium.org/2020/01/ building-more-private-web-path-towards.html. [9] Google. The next step toward phasing out third-party cook- ies in chrome, 2023. https://blog.google/products/chrome/ privacy-sandbox-tracking-protection/. [10] Google. A new path for privacy sandbox on the web, 2024. https:// privacysandbox.com/news/privacy-sandbox-update/. [11] M. Grootendorst. Keybert: Minimal keyword extraction with bert., 2020. [12] H. Guo, R. Tang, Y. Ye, Z. Li, and X. He. Deepfm: A factorization-machine based neural network for ctr prediction, 2017. [13] S. Ioffe and C. Szegedy. Batch normalization: Accelerating deep network training by reducing internal covariate shift, 2015. [14] A. Jain, M. Pathak, and M. Divya Prabha. Tackling cookieless domain recommen- dation for digital advertising targetting. page 111–112, 2022. [15] S.khan,Q.M.Ilyas,andW.Anwar.Contextualadvertisingusingkeywordextraction through collocation. In Proceedings of the 7th International Conference on Frontiers of Information Technology, FIT ’09, New York, NY, USA, 2009. Association for Computing Machinery. [16] D. P. Kingma and J. Ba. Adam: A method for stochastic optimization, 2017. [17] J. Lian, X. Zhou, F. Zhang, Z. Chen, X. Xie, and G. Sun. xdeepfm: Combining explicit and implicit feature interactions for recommender systems. In Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, KDD '18. ACM, July 2018. [18] S. Lloyd. Least squares quantization in pcm. IEEE Transactions on Information Theory, 28(2):129–137, 1982. [19] L. Long. Practice papers effective first-party data collection in a privacy-first world. Applied Marketing Analytics, 7(3):202–210, 2022. [20] MartinThoma. Receiver operating characteristic (roc) curve, 2018. https:// commons.wikimedia.org/w/index.php?curid=70212136. [21] T. Mikolov, K. Chen, G. Corrado, and J. Dean. Efficient estimation of word repre- sentations in vector space, 2013. [22] M. Pachilakis, P. Papadopoulos, E. P. Markatos, and N. Kourtellis. No more chasing waterfalls: A measurement study of the header bidding ad-ecosystem. In Proceedings of the Internet Measurement Conference, IMC ’19, page 280–293, New York, NY, USA, 2019. Association for Computing Machinery. [23] Rakesh4realg. Neighborhood based collaborative filter- ing —part 4, 2019. https://medium.com/fnplus/ neighbourhood-based-collaborative-filtering-4b7caedd2d11. [24] N. Reimers and I. Gurevych. Sentence-bert: Sentence embeddings using siamese bert-networks, 2019. [25] S. Rendle. Factorization machines. In 2010 IEEE International Conference on Data Mining, pages 995–1000, 2010. [26] W. Shen. Deepctr: Easy-to-use,modular and extendible package of deep-learning based ctr models. https://github.com/shenweichen/deepctr, 2017. [27] W.Song,C.Shi,Z.Xiao,Z.Duan,Y.Xu,M.Zhang,andJ.Tang.Autoint:Automatic feature interaction learning via self-attentive neural networks. In Proceedings of the 28th ACM International Conference on Information and Knowledge Management, CIKM '19. ACM, Nov. 2019. [28] R. van Meteren. Using content-based filtering for recommendation. 2000. [29] A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. Kaiser, and I. Polosukhin. Attention is all you need, 2023. [30] J.Wang,W.Zhang,andS.Yuan.Displayadvertisingwithreal-timebidding(rtb)and behavioural targeting. Foundations and Trends in Information Retrieval, 11(4-5):297 –435, 2017. Cited by: 75; All Open Access, Green Open Access. [31] L. Wang, K.-C. Lee, and Q. Lu. Improving advertisement recommendation by en- riching user browser cookie attributes. volume 24-28-October-2016, page 2401 – 2404, 2016. [32] R.Wang,B.Fu,G.Fu,andM.Wang.Deep&crossnetworkforadclickpredictions, 2017. [33] R. Wang, R. Shivanna, D. Cheng, S. Jain, D. Lin, L. Hong, and E. Chi. Dcn v2: Improved deep & cross network and practical lessons for web-scale learning to rank systems. In Proceedings of the Web Conference 2021, WWW '21. ACM, Apr. 2021. [34] Wikipedia.Cosinesimilarity,2007.https://en.wikipedia.org/wiki/Cosine_ similarity. [35] Wikipedia. Kullback–leibler divergence, 2011. https://en.wikipedia.org/ wiki/Kullback%E2%80%93Leibler_divergence. [36] Wikipedia. Elbow method (clustering), 2016. https://en.wikipedia.org/ wiki/Elbow_method_(clustering). [37] Wikipedia. General data protection regulation, 2016. https://en.wikipedia. org/wiki/General_Data_Protection_Regulation. [38] Wikipedia. California consumer privacy act, 2018. https://en.wikipedia.org/ wiki/California_Consumer_Privacy_Act. [39] R. Zhang, Q.-d. Liu, Chun-Gui, J.-X. Wei, and Huiyi-Ma. Collaborative filtering for recommender systems. In 2014 Second International Conference on Advanced Cloud and Big Data, pages 301–308, 2014. [40] Y. Zhang. An introduction to matrix factorization and factorization machines in recommendation system, and beyond, 2022.zh_TW