本論文並使用了自有資料集和公開資料集作為案例應用，開發一個AI ERP系統， 專注於利用資料視覺化、深度學習模型和組合信任鏈，這三大部分做開發。其中 通過資料視覺化技術，將複雜的資料分析結果被轉化為易於理解的圖表和圖形， 有效地提高了信息的可讀性和可用性；另外應用了多種深度學習模型來分析和應 用這些資料，並選擇了LSTM模型進行銷售預測、DNN雙塔模型進行商品推薦和 大語言模型進行無人化客服系統。提供更加直觀和用戶友好的方式來理解和解釋 複雜的資料集，從而為用戶提供有價值的商業洞察；使用組合信任鏈進行全線管 理的保護。平台的成果可以作為大數據之服務平台雛型，最後也期待此平台能提 供使用者在其他領域去在預測銷售趨勢、客戶分析以及輔助決策制定過程中的有 效性，也為未來在此領域的研究提供了寶貴的參考和啟示。

This thesis develops an AI-driven ERP system using both proprietary and open datasets, focusing on three core components: data visualization, deep learning models, and a composite trust chain. Visualization techniques transform complex analytics into intuitive charts, enhancing readability and usability. Deep learning models—LSTM for sales forecasting, a DNN dual-tower model for product recommendation, and an LLM for automated customer service—offer user-friendly insights from complex datasets. The composite trust chain ensures secure end-to-end management.The resulting platform serves as a prototype for big data service applications and demonstrates potential in areas such as sales prediction, customer analysis, and decision support. It also provides valuable guidance for future research in this field.

目錄

致謝 ii 
論文題要 iii 
Abstract iv 
1 簡介 1 
1.1 研究動機與背景 1 
1.2 研究目的 2 
1.3 研究範圍與限制 2 
1.4 研究架構與方法 3 
2 文獻探討 6 
2.1 ERP定義 6 
2.2 Open source雲端、雲端與傳統ERP比較 7 
2.3 ERP成功的關鍵因素 8 
2.4 AI導入ERP的優勢 9 
3 系統實作與開發 10 
3.1 開發環境與工具選擇 10 
3.2 模組化開發與整合 11 
3.3 模組流程介紹 12 
3.4 模組導入流程 16 
3.5 匯入資料與平台功能說明 16 
4 系統分析與評估 38
4.1 功能性指標 38
4.2 可用性指標 39 
4.3 安全性指標 39 
5 研究結論與建議 40
5.1 研究結論 40 
5.2 研究建議 40
參考文獻 42

圖目錄

1.1 平台研究架構 4 
2.1 ERP 定義架構 7 
3.1 Node.js、MySQL和、JavaScript和使用者之間的運作方式 11 
3.2 採購流程圖 13 
3.3 銷售流程圖 14 
3.4 客戶退貨流程圖 15 
3.5 瑕疵品處理流程圖 15 
3.6 系統模組圖 16 
3.7 客戶銷售資訊表 17 
3.8 銷售明細單 17 
3.9 商品留存率圖 19 
3.10 客戶分析圖 20 
3.11 客戶流失率分析熱力圖 20 
3.12 Self-RAG 與用戶交互流程圖 21 
3.13 Ragas評估結果 23 
3.14 年度報表圖 24 
3.15 LSTM預測流程圖 24 
3.16 LSTM模型架構 26 
3.17 銷售明細資料表 28 
3.18 產品與訂單分析圖 28 
3.19 商品推薦系統流程圖 29 
3.20 DNN雙塔模型架構 30 
3.21 推薦結果 31 
3.22 查詢庫存和庫存列表查詢功能 32 
3.23 品項管理清單 33 
3.24 商品資料表格 33 
3.25 採購單示例圖 34 
3.26 採購清單與表格查詢功能 34 
3.27 MFA 與 Session 流程圖 37

表目錄

2.1 Open source 雲端ERP、雲端ERP與傳統ERP系統比較 8 
3.1 留存率演算法 18 
3.2 最大最小正規化符號說明 25 
3.3 指數移動平均符號說明 25 
3.4 模型超參數 27 
3.5 組合信任鏈作法及效果 35 
4.1 Cronbach’s Alpha 符號說明 38 
4.2 Principal Component Analysis 符號說明 39 
4.3 Varimax 旋轉方法符號說明 39 

[1] 黃紓芸, "規劃 ERP 系統設計與查詢策略之實踐 - 以某中小企業為例," in 明新 
科技大學資訊管理系學位論文, pp. 1–98, 2023.  
[2] 林育震, "掌控風險發揮雲端效益," in Chinese Cryptology and Information  
Security Association, vol. 16, no. 4, pp. 138–149, 2010. 
[3] 醉翁, "推薦系統-雙塔模型基本原理與實踐," zhihu.com, [Online.] Available:  
https://zhuanlan.zhihu.com/p/601743047, [Accessed: Jan. 31, 2023]. 
[4] 陳興源, "自行車零件銷售預測 - 以N公司為例," in 國立雲林科技大學資訊管理 
系學位論文, pp. 1–45, 2023. 
[5] 陳泓伸, "中小企業導入Open Source雲端ERP的效益分析 - 以WebERP為例,"  
in 逢甲大學工業工程與系統管理學系學位論文, pp. 1–93, 2017. 
[6] 黃之柔, "用 LSTM 神經網路預測銷售額-以小型企業為例," in 國立成功大學 
工程科學系學位論文, pp. 1–66, 2023. 
[7] 蔡文賢, 簡世文, 呂俊德, 許秉瑜, 范懿文, 與 鄭明松, "建構ERP系統之績效評 
估指標," in 電子商務學報, vol. 5, no. 1, pp. 107–134, 2023. 
[8] C. Li, "ERP package: What’s Next?," in Handbook of Data Management, 1999 ed., 
London: Auerbach Publications, 1999, pp. 8. 
[9]  F. Liu, Z. Kang, X. Han. "Optimizing RAG Techniques for Automotive Industry  
PDF Chatbots: A Case Study with Locally Deployed Ollama Models," in arXiv 
preprint arXiv:2408.05933, 2024. 
[10] M. Armbrust, A. Fox, R. Griffith, A. D. Joseph, R. Katz, A. Konwinski, et al., "A 
view of cloud computing," in Commun. ACM, vol. 53, no. 4, pp. 50–58, 2010. 
[11] F. F.-H. Nah, J. L.-S. Lau, and J. Kuang, "Critical Factors for Successful 
Implementation of Enterprise Systems," in Bus. Process Manag. J., vol. 7, no. 3, 
pp. 285–296, 2001. 
[12] J. Ram, D. Corkindale, and M.-L. Wu, "Implementation Critical Success Factors 
(CSFs) for ERP: Do they Contribute to Implementation Success and Post- 
Implementation Performance," in Int. J. Prod. Econ.,vol. 144, no. 1, pp. 157–174,  
2013.    
[13] F. Chollet, Deep Learning with Python. New York, NY, USA: Manning, 2021. 
[14] A. Asai, Z. Wu, Y. Wang, A. Sil, and H. Hajishirzi, "Self-RAG: Learning to 
Retrieve, Generate, and Critique through Self-Reflection," in Proc. 12th Int. Conf.  
Learning Representations (ICLR), 2024. 
[15] S. S. Bawa, "How Business can use ERP and AI to become Intelligent Enterprise,"  
Int. J. Innovative Sci. Res. Technol., vol. 8, pp. 8–11, 2023. 
[16] H. Jiang, J. Ruan, and J. Sun, "Application of Machine Learning Model and Hybrid  
Model in Retail Sales Forecast," in 2021 IEEE 6th Int. Conf. Big Data Analytics  
(ICBDA), pp. 69–75, 2021 
[17] M. Alawneh and I. M. Abbadi, "Integrating Trusted Computing Mechanisms with 
Trust Models to Achieve Zero Trust Principles," in 2022 9th Int. Conf. Internet  
Things: Syst., Manage. Security (IOTSMS), pp. 1–6, 2022. 
[18] S. Es, J. James, L. Espinosa-Anke and S. Schockaert, "RAGAS:Automated 
Evaluation of Retrieval Augmented Generation," in Proc. 18th Conf. Eur. Chapter 
Assoc. Comput. Linguistics: Syst. Demonstrations, pp. 150–158, 2024. 
[19] Z. Nussbaum, J. X. Morris, B. Duderstadt, and A. Mulyar, "Nomic Embed: Training 
a Reproducible Long Context Text Embedder," arXiv preprint arXiv:2402.01613, 
2024. 
[20] S. Xiao, Z. Liu, P. Zhang, and N. Muennighoff, "C-Pack: Packaged Resources to 
Advance General Chinese Embedding," arXiv preprint arXiv:2309.07597, 2023. 
[21] A. Grattafiori et al., "The LLaMA 3 herd of models," arXiv preprinarXiv:2407.21783, 
2024. 
[22] P. Rajpurkar, J. Zhang, K. Lopyrev, and P. Liang, "SQuAD: 100,000+ questions for 
machine comprehension of text," in Proc. 2016 Conf. Empirical Methods Natural 
Language Processing (EMNLP), pp. 2383–2392, 2016.