本研究以 A 化學工業公司導入作業價值管理（AVM）為個案，探討如何運用文字分類技術支援模組一資源分類與模組三作業分類，並評估 AVM 模組資料智慧化導入支援之效益。研究以既有財務與工作職掌資料為基礎，經文字前處理與階層標籤整理後，比較 Naïve Bayes、Logistic Regression 與 Decision Tree 之分類表現，並結合模型建議、信心分數、理由線索、候選清單與分流規則，建構人機協作覆核流程。
研究結果顯示，AVM 模組資料智慧化導入支援可將部分逐筆人工判斷轉換為模型輔助下之聚焦覆核；依本研究設定之相同任務範圍與推估基礎，人工流程約需 2,708 分鐘，智慧化流程約需 1,157 分鐘，推估可節省約 1,551 分鐘，整體時間節省率約為 57.27%。本研究認為，文字分類技術雖不能取代顧問專業判斷，但可作為 AVM 導入分類整理與覆核之輔助工具，支持分類判斷之一致性、流程可追溯性與導入效率。

This study uses the implementation of Activity Value Management (AVM) at A Chemical Industrial Company as a case to examine how text classification techniques can support Module 1 resource classification and Module 3 activity classification, and to evaluate the benefits of intelligent implementation support for AVM model data. In this study, AVM model data refers to the foundational classification data required for AVM module construction before formal AVM computation. Based on existing financial and job responsibility data, this study performs text preprocessing and hierarchical label organization, compares the classification performance of Naïve Bayes, Logistic Regression, and Decision Tree, and integrates model recommendations, confidence scores, rationale cues, candidate lists, and routing rules to construct a human-machine collaborative review process.
The results show that the intelligent support process for AVM model data can transform part of the item-by-item manual judgment into model-assisted focused review. Based on the task scopes and estimation assumptions defined in this study, the manual process requires approximately 2,708 minutes, while the intelligent process requires approximately 1,157 minutes, resulting in a time saving rate of approximately 57.27%. This study argues that although text classification techniques cannot replace consultants’ professional judgment, they can serve as a supporting tool for classification organization and review before formal AVM computation, thereby supporting consistency in classification judgments, process traceability, and implementation efficiency.

一、 中文文獻部分
吳安妮，2007，確立管理方向設計專屬ABC－作業基礎成本制之發展與整合，會計研究月刊，第 263 期：60-74。
吳安妮，2015，管理會計技術商品化：以ABC為核心之作業價值管理系統（AVMS）為例，會計研究月刊，第 359 期：20-24。
吳安妮，2021，企業策略的終極答案：用「作業價值管理 AVM」破除成本迷思，掌握正確因果資訊，做對決策賺到「管理財」，台北：臉譜出版。
李紹瑄與黃政仁，2026，結合人工智慧之產品生命週期價值與作業價值管理研究：以 P 通路代理商為例，臺大管理論叢，第 36 卷，第 1 期：43-106。
劉順仁、王俊凱、李季澄與黃禹翔，2021，機器學習與會計科目判斷－台灣半導體公司使用人工智慧提升行政效率個案對會計教育之啟示，中華會計學刊，第 17 卷，第 2 期：245-268。
二、 英文文獻部分
Abbas, K. 2026. Management accounting and artificial intelligence: A comprehensive literature review and recommendations for future research. The British Accounting Review 58: 101551.
Bodendorf, F., and J. Franke. 2024. Synthesis of activity-based costing and deep learning to support cost management: A case study in the automotive industry. Computers & Industrial Engineering 196: 110449.
Cooper, R., and R. S. Kaplan. 1988. Measure costs right: Make the right decisions. Harvard Business Review 66 (5): 96-103.
Hüllermeier, E., and W. Waegeman. 2021. Aleatoric and epistemic uncertainty in machine learning: An introduction to concepts and methods. Machine Learning 110: 457-506.
Jørgensen, M., and M. Shepperd. 2007. A systematic review of software development cost estimation studies. IEEE Transactions on Software Engineering 33 (1): 33-53.
Kaplan, R. S., and S. R. Anderson. 2004. Time-driven activity-based costing. Harvard Business Review 82 (11): 131-138.
Kaplan, R. S., and D. P. Norton. 1992. The balanced scorecard—Measures that drive performance. Harvard Business Review 70 (1): 71-79.
Knox, B. D. 2023. Machine learning activity-based costing: Can activity-based costing’s first-stage allocation be replaced with a neural network? Journal of Emerging Technologies in Accounting 20 (1): 1-17.
Lai, V., and C. Tan. 2019. On human predictions with explanations and predictions of machine learning models: A case study on deception detection. Proceedings of the Conference on Fairness, Accountability, and Transparency: 29-38.
Lebovitz, S., H. Lifshitz-Assaf, and N. Levina. 2022. To engage or not to engage with AI for critical judgments: How professionals deal with opacity when using AI for medical diagnosis. Organization Science 33 (1): 126-148.
Manning, C. D., P. Raghavan, and H. Schütze. 2009. An introduction to information retrieval. Cambridge: Cambridge University Press.
Mozannar, H., and D. Sontag. 2020. Consistent estimators for learning to defer to an expert. Proceedings of the 37th International Conference on Machine Learning: 7076-7087.
Raschka, S., and V. Mirjalili. 2019. Python machine learning: Machine learning and deep learning with Python, scikit-learn, and TensorFlow 2 (3rd ed.). Birmingham: Packt Publishing.