跳到主要內容

臺灣博碩士論文加值系統

(18.97.14.91) 您好!臺灣時間:2025/02/19 19:56
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

我願授權國圖
: 
twitterline
研究生:王健嘉
研究生(外文):WANG, CHIEN-CHIA
論文名稱:節能專利領域運用到 AI 演算法之趨勢研究分析
論文名稱(外文):Research and analysis on the application trend of artificial intelligence algorithms in the field of energy-saving patents
指導教授:李達生李達生引用關係
指導教授(外文):LEE, DA-SHENG
口試委員:陳清祺陳廷昱
口試委員(外文):CHENG, CHIN-CHICHEN, TING-YU
口試日期:2022-07-14
學位類別:碩士
校院名稱:國立臺北科技大學
系所名稱:能源與冷凍空調工程系
學門:工程學門
學類:其他工程學類
論文種類:學術論文
論文出版年:2022
畢業學年度:110
語文別:中文
論文頁數:63
中文關鍵詞:節能專利AI 演算法自然語言處理主題模型關鍵字趨勢分析
外文關鍵詞:Energy-Saving PatentsAI AlgorithmsNatural Language ProcessingTopic ModelKeywordTrend Analysis
相關次數:
  • 被引用被引用:0
  • 點閱點閱:322
  • 評分評分:
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:1
Artificial intelligence(AI)隨著數據運算速度的提升,發展出許多演算方法。而在節能技術中用到的數據運算,亦發展出各項AI演算法應用於不同節能場域。其中,在節能專利技術開發上,更能顯示出AI演算法對於能源利用的價值與技術發明的獨創性。因此,為了評估AI演算法用於節能技術是否有其適用性與發展性,本研究從2014~2021年度之間超過15,000篇節能專利中,對於AI演算法與節能領域進行趨勢分析。
透過撰寫Python程式以及自然語言處理Natural Language Processing(NLP)邏輯,使用隱含狄利克雷分佈(LDA)做為建立主題模型的方法,將大量節能專利資料進行主題分類,並且輔以AI演算法關鍵字做為節能專利的標籤,用來加強AI演算法應用於節能領域的相關性。歸納這些AI演算法以及各種領域的節能專利後找出發展趨勢,說明了節能專利使用到的各種AI演算法其逐年間的消長趨勢做探討分析。進而統整出AI演算法對應於節能場域的適性,做技術應用的探討,以作為AI演算法應用於節能專利之開發參考。
經過分析所有節能專利後,可以歸納出用於節能專利的AI演算法趨勢,早期以神經網絡Neural Network(NN)、機器學習Machine Learning(ML)、基因演算法Genetic Algorithm(GA)為主,近年隨著各項技術與應用的發展,深度學習Deep Learning(DL)與強化學習Reinforce learning(RL)在應用數量上有大幅增長。在近年疫情的催化下,人們行為模式的改變為遠距離交流、居家辦公等,對於能源的需求反映在能源優化、通訊網路以及電動車相關的主題,並延伸出智慧工廠、智慧家居、智慧醫療和綠能環保等主題內容,可以預見以AI為主軸的應用,會圍繞在日常生活之中。
Artificial intelligence (AI) has developed many calculation methods with the improvement of data processing speed. For the data operations used in energy-saving technologies, various AI algorithms have also been developed to apply to different energy-saving fields. Among them, in the development of energy-saving patented technology, it can better show the value of AI algorithm for energy utilization and the originality of technological invention. Therefore, in order to evaluate the applicability and development of AI algorithms for energy-saving technologies, this research conducted a trend analysis of AI algorithms and energy-saving fields from more than 15,000 energy-saving patents between 2014 and 2021.
By writing Python programs and Natural Language Processing (NLP) logic, using latent Dirichlet distribution (LDA) as a method to establish topic models, a large number of energy-saving patent data are classified by topic, and AI algorithm is used to provide the keywords is used as a label for energy-saving patents to strengthen the relevance of AI algorithms in the field of energy-saving. After summarizing these AI algorithms and energy-saving patents in various fields, we can find out the development trend, and discuss and analyze the trend of the rise and fall of various AI algorithms used in energy-saving patents year by year. Then, the suitability of the AI algorithm for the energy-saving field is unified, and the technical application is discussed, which can be used as a reference for the application of the AI algorithm to the development of energy-saving patents.
After analyzing all energy-saving patents, the trend of AI algorithms for energy-saving patents can be summarized. In the early days, neural network Neural Network (NN), Machine Learning (ML), and Genetic Algorithm (GA) were the main ones. With the development of various technologies and applications, the number of applications of Deep Learning (DL) and Reinforcement Learning (RL) has increased significantly. After over two years of living with Covid-19, it is not only employees shifted to remote work but also consumers are changing their behavior during the crisis. More and more emphasis being placed on the Energy-related topics such as energy optimization, ICT, electric vehicles, smart manufacturing, home automation, smart healthcare and green/renewable energy etc. It is foreseeable that the application of AI as the main axis will revolve around daily life.
摘要 i
ABSTRACT iii
誌謝 v
目錄 vi
表目錄 ix
圖目錄 x
第一章 導論 1
1.1 前言 1
1.2 研究動機與目的 2
1.3 文獻探討 3
1.3.1 自然語言概述 3
1.3.2 主題模型 (Topic model) 5
1.3.3 趨勢分析 7
第二章 研究理論 8
2.1 程式語言與自然語言 8
2.1.1 機器學習 8
2.1.2 深度學習 9
2.1.3 自然語言處理(NLP) 9
2.2 隱含狄利克雷分布(LDA) 10
2.2.1 LDA數學模型 10
2.2.2 吉布斯取樣於LDA參數估計 11
2.3 Python 13
2.3.1 Jupyter Notebook 13
2.3.2 Gensim 13
2.3.3 nltk 13
2.4 全球專利檢索系統(GPSS系統) 14
第三章 研究設計與實施 15
3.1 資料收集 16
3.2 資料前處理 16
3.3 以AI演算法為分類的主題標籤 18
3.3.1 定義主要關鍵主題詞庫 19
3.3.2 提取次要候選關鍵主題詞 19
3.3.3 權重分數 20
3.3.4 AI演算法主題分類標籤 20
3.4 LDA主題模型 21
3.4.1 字典和語料庫 22
3.4.2 基本模型 22
3.4.3 連貫性分數評估模型主題數量 23
3.4.4 調整超參數 24
3.4.5 建立最終LDA主題模型 24
第四章 研究分析與結果討論 26
4.1 年度AI演算法分析 27
4.1.1 各年度使用於專利的AI演算法分類結果 27
4.1.2 總年度AI演算法統計 32
4.1.3 2014~2021總年度AI演算法數量統計分析 34
4.2 利用LDA模型將節能專利應用的主題分類 35
4.2.1 各年度專利主題分類 35
4.2.2 專利主題領域分類與趨勢 52
第五章 結論與建議 58
5.1 結論 58
5.2 未來展望與建議 60
參考文獻 62
[1] J. Lee and J. Hsiang, “Patent Classification by Fine-tuning BERT Language Model,” World Patent Information, vol. 61, pp.1-4, Jun. 2020.
[2] M. Sofean, “Deep Learning Based Pipeline with Multichannel Inputs for Patent Classification, World Patent Information, vol. 66, pp. 1-6, Sep. 2021.
[3] E. Saari,“Trend Analysis in AI Research Over Time Using NLP Techniques,” Tampere University, 2019.
[4] J. Devlin, M. Chang, K. Lee and K. Toutanova, “BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding,”arXiv, preprint arXiv: 1810.04805, 2018.
[5] 陳韻卉(2020)。利用主題模型擷取跨語言專利和趨勢分析之研究。國立中山大學資訊管理學系研究所碩士論文,高雄市。
[6] R. Gupta, R. Agarwalla, B. Naik, J. Evuri, A. Thapa and T. Singh,“Prediction of Research Trends Using LDA Based Topic Modeling,”Global Transitions Proceedings, vol. 3, no. 1, pp. 298-304, Jun. 2022.
[7] N. Yang, J. Jo, M. Jeon, W. Kim and J. Kang,“Semantic and explainable Research-related Recommendation System Based on Semi-supervised Methodology Using BERT and LDA Models,”Expert Systems with Applications, vol. 190, p. 116209, Mar. 2022.
[8] Q. Xie, X. Zhang, Y. Ding and M. Song, “Monolingual and Multilingual Topic Analysis Using LDA and BERT Embeddings,”Journal of Informetrics, vol. 14, no. 3, p. 101055, Aug. 2020.
[9] E. Zvornicanin,“When Coherence Score is Good or Bad in Topic Modeling?,”Available: https://www.baeldung.com/cs/topic-modeling-coherence-score (accessed on May 15 2021).
[10] J.L. Aleixandre-Tud, L.C. Castelló-Cogollosbc, J.L. Aleixandre and R. Aleixandre-Benaventce “Renewable Energies: Worldwide Trends in Research, Funding and International Collaboration,” Renewable Energy, vol. 139, pp. 268-278, Aug. 2019.
[11] N. Zhanga, Y. Choi and W. Wang, “Does Energy Research Funding Work? Evidence From The Natural Science Foundation of China Using TEI@I Method,”Technological Forecasting and Social Change, vol. 144, pp. 369-380, Jul. 2019.
[12] M. Du, B. Wang and N. Zhang, “National Research Funding and Energy Efficiency: Evidence From The National Science Foundation of China,” Energy Policy, vol. 120, pp. 335-346, Sep. 2018.
[13] 魏明昇, “利用專利分析探討太陽能熱水系統之發展趨勢研究,” 崑山科技大學, 2019.
[14] 中央研究院,“斷開中文的鎖鍊!自然語言處理 (NLP),”Available: https://research.sinica.edu.tw/nlp-natural-language-processing-chinese-knowledge-information/ (accessed on July 03 2018).
[15] D.M. Blei, A.Y. Ng and M.I. Jordan,“Latent Dirichlet Allocation,”Journal of Machine Learning Research, vol. 3, pp. 993-1022, Mar. 2003.
[16] 維基百科,“隱含狄利克雷分布,”Available: https://zh.wikipedia.org/zh-tw/%E9%9A%90%E5%90%AB%E7%8B%84%E5%88%A9%E5%85%8B%E9%9B%B7%E5%88%86%E5%B8%83 (accessed on Apr. 09 2021).
[17] 經濟部智慧財產局,“中華民國專利資訊檢索系統 -- 經濟部智慧財產局, Available: https://iknow.stpi.narl.org.tw/post/patentsearch.aspx?&Area_No=11 (accessed on Jul. 01 2008).

電子全文 電子全文(網際網路公開日期:20261122)
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
無相關期刊