研究生: |
翁子平 Weng, Zi-Ping |
---|---|
論文名稱: |
基於注意力機制之高精確知識融合方法之設計-以酪農知識問答系統建置為例 Design of a High Accuracy Knowledge Fusion Scheme Based on Attention Mechanism–Implementation of Diary Knowledge QA System |
指導教授: |
龔旭陽
Kung, Hsu-Yang |
學位類別: |
碩士 Master |
系所名稱: |
管理學院 - 資訊管理系所 Department of Management Information Systems |
畢業學年度: | 109 |
語文別: | 中文 |
論文頁數: | 66 |
中文關鍵詞: | 知識問答系統 、知識圖譜 、注意力機制模型 、知識分歧 、知識融合 、知識抽取 、TF-IDF 、雙向門控循環單元 、條件隨機域 、酪農知識 |
外文關鍵詞: | Knowledge Question Answering, Attention Mechanism Model, Knowledge Divergence, Knowledge Fusion, Term frequency-Inverse document frequency, Bidirectional Gate Recurrent Unit, Conditional Random Field, Dairy Farmer Knowledge |
DOI URL: | http://doi.org/10.6346/NPUST202100343 |
相關次數: | 點閱:24 下載:0 |
分享至: |
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 |
摘 要:I
謝 誌:IX
圖目錄:XIV
表目錄:XVII
1. 緒論:1
1.1 研究背景與動機:1
1.2 研究目的與方法:3
1.3 論文架構:5
2. 文獻探討與技術背景:6
2.1 資料前處理:6
2.1.1 TF-IDF (Term frequency–Inverse document frequency):6
2.1.2 隱藏馬可夫模型(Hidden Markov Model, HMM):6
2.2 知識三元組:7
2.2.1 分佈式表示(Distributional Representation):7
2.2.2 詞嵌入(word embedding):8
2.2.3 資源描述框架(Resource Description Framework, RDF):9
2.2.4 命名實體識別(Named Entity Recognition, NER):9
2.2.5 雙向門控循環單元(Bidirectional Gated Recurrent Unit, Bi-GRU):10
2.2.6 條件隨機域(Conditional Random Field, CRF):11
2.3 知識建構(Knowledge Construction):12
2.3.1 三元組抽取:12
2.3.2 知識融合:13
2.3.3 注意力機制模型(Attention-based Model):14
2.4 知識檢索與知識問答:14
2.4.1 問句預處理:14
2.4.2 主要實體探索:15
2.4.3 Cypher:15
2.5 酪農產業現況:15
3. 研究方法與步驟:17
3.1 研究問題定義:17
3.2 系統架構:17
3.3 系統功能說明:19
3.4 知識抽取層:20
3.4.1 資料前處理:22
3.4.2 知識三元組:25
3.5 知識建構層:29
3.5.1 知識圖譜建置:31
3.5.2 知識驗證:33
3.6 知識檢索層:33
3.6.1 問句預處理:34
3.6.2 關聯檢索:35
3.7 知識表示層:37
3.7.1 候選答案建立:38
3.7.2 知識問答系統:40
4. 系統實作與效能評估:41
4.1 系統實作與效能評估:41
4.2 實驗硬體:41
4.3 開發軟體:41
4.4 資料集:42
4.5 知識抽取層:43
4.5.1 資料前處理:43
4.5.2 知識三元組:47
4.6 知識建構層:53
4.6.1 知識融合:53
4.6.2 注意力機制之知識融合:56
4.7 知識檢索層:57
4.7.1 問句預處理:57
4.7.2 關聯檢索:58
4.8 知識檢索層:59
4.8.1 候選答案排名:59
4.8.2 知識問答系統:61
5. 結論:63
6. 參考文獻:65
作者簡介:68
[1]國立台灣大學,文字探勘之前處理與TF-IDF介紹,取自:http://www.cc.ntu.edu.tw/chinese/epaper/0031/20141220_3103.html
[2]呂國豪,(2018),結合NLP與Bi-LSTM之智慧飼養知識計算系統設計與實作。
[3]行政院農委會,農業統計資料,取自:https://agrstat.coa.gov.tw/sdweb/public/inquiry/InquireAdvance.aspx
[4]行政院農委會,農政農情2020年1月,取自:https://www.coa.gov.tw/ws.php?id=2509988
[5]Arnaout, H., & Elbassuoni, S. (2018). Effective searching of RDF knowledge graphs. Journal of Web Semantics, 48, 66-84.
[6]Bafna, P., Pramod, D., & Vaidya, A. (2016, March). Document clustering: TF-IDF approach. In 2016 International Conference on Electrical, Electronics, and Optimization Techniques (ICEEOT) (pp. 61-66). IEEE.
[7]Choi, H., Cho, K., & Bengio, Y. (2018). Fine-grained attention mechanism for neural machine translation. Neurocomputing, 284, 171-176.
[8]Chorowski, J., Bahdanau, D., Serdyuk, D., Cho, K., & Bengio, Y. (2015). Attention-based models for speech recognition. arXiv preprint arXiv:1506.07503.
[9]Cho, K., Van Merriënboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H., & Bengio, Y. (2014). Learning phrase representations using RNN encoder-decoder for statistical machine translation. arXiv preprint arXiv:1406.1078.
[10]Harris, Z. S. (1954). Distributional structure. Word, 10(2-3), 146-162.
[11]Heim, P., Hellmann, S., Lehmann, J., Lohmann, S., & Stegemann, T. (2009, December). RelFinder: Revealing relationships in RDF knowledge bases. In International Conference on Semantic and Digital Media Technologies (pp. 182-187). Springer, Berlin, Heidelberg.
[12]Kim, S. W., & Gil, J. M. (2019). Research paper classification systems based on TF-IDF and LDA schemes. Human-centric Computing and Information Sciences, 9(1), 1-21.
[13]Kaiser, Ł., & Sutskever, I. (2015). Neural gpus learn algorithms. arXiv preprint arXiv:1511.08228.
[14]Lenci, A. (2008). Distributional semantics in linguistic and cognitive research. Italian journal of linguistics, 20(1), 1-31.
[15]Lafferty, J., McCallum, A., & Pereira, F. C. (2001). Conditional random fields: Probabilistic models for segmenting and labeling sequence data.
[16]Luo, L., Yang, Z., Yang, P., Zhang, Y., Wang, L., Lin, H., & Wang, J. (2018). An attention-based BiLSTM-CRF approach to document-level chemical named entity recognition. Bioinformatics, 34(8), 1381-1388.
[17]Mikolov, T., Sutskever, I., Chen, K., Corrado, G., & Dean, J. (2013). Distributed representations of words and phrases and their compositionality. arXiv preprint arXiv:1310.4546.
[18]Miller, E. (1998). An introduction to the resource description framework. Bulletin of the American Society for Information Science and Technology, 25(1), 15-19.
[19]Nadeau, D., & Sekine, S. (2007). A survey of named entity recognition and classification. Lingvisticae Investigationes, 30(1), 3-26.
[20]Francis, N., Green, A., Guagliardo, P., Libkin, L., Lindaaker, T., Marsault, V., ... & Taylor, A. (2018, May). Cypher: An evolving query language for property graphs. In Proceedings of the 2018 International Conference on Management of Data (pp. 1433-1445).
[21]Qaiser, S., & Ali, R. (2018). Text mining: use of TF-IDF to examine the relevance of words to documents. International Journal of Computer Applications, 181(1), 25-29.
[22]Ren, S., & Li, H. (2020, February). Question Answering Model Based on Graph Knowledge and Entity Recognition. In 2020 12th International Conference on Measuring Technology and Mechatronics Automation (ICMTMA) (pp. 578-581). IEEE.
[23]Wang, F., Jiang, M., Qian, C., Yang, S., Li, C., Zhang, H., ... & Tang, X. (2017). Residual attention network for image classification. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 3156-3164).
[24]Zhou, G., Xie, Z., Yu, Z., & Huang, J. X. (2021). DFM: A parameter-shared deep fused model for knowledge base question answering. Information Sciences, 547, 103-118.