| dc.contributor.advisor | 鄭宇庭 | zh_TW |
| dc.contributor.author (Authors) | 周正娟 | zh_TW |
| dc.contributor.author (Authors) | Zhou, Zheng-Juan | en_US |
| dc.creator (作者) | 周正娟 | zh_TW |
| dc.creator (作者) | Zhou, Zheng-Juan | en_US |
| dc.date (日期) | 2021 | en_US |
| dc.date.accessioned | 4-Aug-2021 16:38:30 (UTC+8) | - |
| dc.date.available | 4-Aug-2021 16:38:30 (UTC+8) | - |
| dc.date.issued (上傳時間) | 4-Aug-2021 16:38:30 (UTC+8) | - |
| dc.identifier (Other Identifiers) | G0108363084 | en_US |
| dc.identifier.uri (URI) | http://nccur.lib.nccu.edu.tw/handle/140.119/136733 | - |
| dc.description (描述) | 碩士 | zh_TW |
| dc.description (描述) | 國立政治大學 | zh_TW |
| dc.description (描述) | 企業管理研究所(MBA學位學程) | zh_TW |
| dc.description (描述) | 108363084 | zh_TW |
| dc.description.abstract (摘要) | 2020 年受到新冠疫情的影響，民眾的生活習慣出現了大幅的改變。為避免 感染病毒，消費者漸漸降低去實體通路的頻率，轉而在電商平台購買生活必需 品。根據財政部公布的電子發票資料顯示，主要電商平台於 2 月份的銷售金額 達到 138.5 億元新台幣，較去年同期成長約 4 成(劉麗惠，2020)。然而隨著更 多民眾開始在網路上購買商品，網路購物假貨氾濫的問題也日益嚴重。根據行 政院消保處所公布之「電商消費爭議總申訴量」資料，六大電商，分別為蝦皮 購物、momo 購物、PChome、露天市集、Yahoo 奇摩購物中心與東森易購，於 2020 年合計獲得 4,243 件的消費者申訴，較去年成長 45.2%。其中，作為 C2C 交易平台的「蝦皮購物」獲得高達 2,029 件申訴，相較去年增幅達到 65%，幾 乎佔去了總申訴案件的一半(陳冠榮，2021)。由於 C2C 交易平台的性質，允許身為一般顧客我們也能夠在該平台開設商 店並販售商品。因此，在賣家人數眾多的情況下，C2C 電商平台較難一一瞭解 並控管個別賣家貨源狀況。因此，本次研究將蒐集蝦皮「美妝保健」分類中， 多位賣家所獲得之顧客評論，並以此建立「偵測假貨」之情感分析模型。希望 未來能將該模型應用於 C2C 電商平台，降低未來顧客受騙上當之事件。 | zh_TW |
| dc.description.abstract (摘要) | Under the influence of Covid-19, consumer behavior changes dramatically. Nowadays, consumers tend to buy daily necessities online in order to prevent spread Covid-19. According to the data released by ministry of finance, the total revenue of the major E-commerce brands was estimated to reach 13.8 billion New Taiwan dollars in February, 2020. The total revenue of the major E-commerce brands was estimated to increase by 40% compared to last year. However, as more and more people start to go online shopping, it also fueled a rapid increase in the trade of counterfeit products on the e-commerce. According to the data released by department of consumer protection, there were about 4,243 E-commerce disputes between sellers and buyers in 2020. The number of disputes increased by 65% compared the last year. Among all the major E- commerce brands, Shopee was accounted for more than half of the disputes.As a C2C e-commerce, Shopee allows everyone to open their own stores on its platform, which appealed to a great number of sellers to sell products on their platform. However, having so many seller selling products on their platform also causes lots of trouble to Shopee when it comes to checking the source credibility of the products sold on its platform. Hence, in this research, I will collect consumer reviews from sellers in the “beauty” category and use the reviews to build sentiment analysis model which was built to detect potential counterfeit products. This research aims to solve the problem of counterfeit products on C2C E-commerce with this sentiment analysis model. | en_US |
| dc.description.tableofcontents | 謝誌 2摘要 3Abstract 4目錄 5表目錄 6圖目錄 7第一章 緒論 9第一節、研究動機 9第二節、研究目的 13第三節、 研究流程 15第二章、文獻探討 16第一節、自然語言處理相關研究 16第二節、情感分析相關研究 21第三章、研究方法 25第一節、研究架構 26第二節、 語料庫訓練 26第三節、訓練資料處理 29第四節、建立 LSTM 情感模型 31第四章、實驗結果與分析 34第一節、Shopee 電商平台 34第二節、建模前置作業結果 37第三節、 LSTM 建模結果 42第五章、結論與建議 45第一節、「貨源可信度」指標計算方式 45第二節、針對蝦皮購物現有作法所提之建議 46第六章、參考文獻 49 | zh_TW |
| dc.format.extent | 5977457 bytes | - |
| dc.format.mimetype | application/pdf | - |
| dc.source.uri (資料來源) | http://thesis.lib.nccu.edu.tw/record/#G0108363084 | en_US |
| dc.subject (關鍵詞) | C2C電商 | zh_TW |
| dc.subject (關鍵詞) | 蝦皮購物 | zh_TW |
| dc.subject (關鍵詞) | 情感分析模型 | zh_TW |
| dc.subject (關鍵詞) | 自然語言處理 | zh_TW |
| dc.subject (關鍵詞) | 假貨偵測 | zh_TW |
| dc.subject (關鍵詞) | C2C E-commerce | en_US |
| dc.subject (關鍵詞) | Sentiment analysis | en_US |
| dc.subject (關鍵詞) | Natural language process | en_US |
| dc.subject (關鍵詞) | Shopee | en_US |
| dc.subject (關鍵詞) | Detecting counterfeit products | en_US |
| dc.title (題名) | 應用情感分析於電商評論偵測潛在仿冒品 ──以 Shopee電商為例 | zh_TW |
| dc.title (題名) | Application of Sentiment Analysis Technology in E-commerce Platform to Detect Potential Counterfeit | en_US |
| dc.type (資料類型) | thesis | en_US |
| dc.relation.reference (參考文獻) | 一、 中文文獻:1. 劉麗惠(2020)。掌握風險迎接新亮點:產業風波，一隻病毒肆虐兩樣情。貿易雜誌，NO.346。2021 年 5 月 29 日，取自https://www.ieatpe.org.tw/magazine/ebook346/storypage02.html2. 程倚華(2020)。momo 成長 3 成、PChome 營收破 200 億!電商雙雄上半年靠疫情受惠多少?。數位時代。2021 年 5 月 29 日，取自 https://www.bnext.com.tw/article/58847/first-half-of-taiwan- e-commerce3. 經濟部智慧財產局(2019)。立法院三讀通過專利法部分條文修正案 及著作權法第 87 條、第 93 條修正案。智慧財產權電子報，NO.155。 2021年5月29日，取自 https://pcm.tipo.gov.tw/pcm2010/pcm/news2_detail.aspx?id=3104. 匠心志(2018)。Guccic 害怕被山寨，放棄與阿里、京東合作。每日 頭條。2021 年 5 月 29 日，取自 https://kknews.cc/zh-tw/tech/6zmgp3q.html5. 陳建鈞(2019)。Nike 斷然分手亞馬遜!撤離全球最大電商平台，背 後 2 大關鍵因素是什麼?。數位時代。2021 年 5 月 29 日，取自 https://www.bnext.com.tw/article/55488/nike-departure-amazon6. 陳冠榮(2021)。國內電商消費糾紛案件大增，蝦皮購物逾 2,000 件 居冠。科技新報。2021 年 5 月 29 日，取自 https://technews.tw/2021/02/08/2020-consumer-disputes-over- online-shopping-platform-in-taiwan/7. Yi-Lun Wu(2019)，多語語碼轉換之未知詞擷取，中華民國計算語言學學會8. 林千翔、張嘉惠、陳貞伶(2010)，結合長詞優先與序列標記之中文 斷詞研究，國立中央大學。二、英文文獻1. Jurafsky,D.;Martin,J.H. (2005). Speech and Language Processing: An Introduction to Natural Language Processing, Computational Linguistic and Speech Recognition. Pearson College Div.2. Pei-Jung, Tsai. (2019). Training a Machine to Read, Comprehend and Answer Questions using Deep Learning. National Chung Cheng University.3. Keith D. Foote, (2019). A Brief History of Natural Language Processing. Dataversity.4. Wong, K.F., W., Xu, R., & Zhang, Z. S. (2009). Introduction to Chinese natural language processing. Synthesis Lectures on Human Language Technologies, 2(1), 1-148.5. L.Jin, (2013). Number in Chinese: A Corpus-Based Computational Investigation.6. Yu, Y., Si, X., Hu, C., & Zhang, J. (2019). A review of recurrent neural networks: LSTM cells and network architectures. Neural computation, 31(7), 1235-1270.7. Sherstinsky, A. (2020). Fundamentals of recurrent neural network (RNN) and long short-term memory (LSTM) network. Physica D: Nonlinear Phenomena, 404, 132306.8. Liu, B. (2012). Sentiment analysis and opinion mining. Synthesis lectures on human language technologies, 5(1), 1-167.9. Liu P, Joty S, Meng H. (2015). Fine-grained opinion mining with recurrent neural networks and word embeddings[C]. Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing.10. Velikovich L, Blair-Goldensohn S, Hannan K, et al. (2010). The viability ofweb-derived polarity lexicons[C].Human Language Technologies: The 2010 Annual Conference of the North American Chapter of the Association for Computational Linguistics. | zh_TW |
| dc.identifier.doi (DOI) | 10.6814/NCCU202100726 | en_US |
