| DC 欄位 |
值 |
語言 |
| DC.contributor | 資訊管理學系 | zh_TW |
| DC.creator | 蔡亞真 | zh_TW |
| DC.creator | Ya-Jen Tsai | en_US |
| dc.date.accessioned | 2024-7-17T07:39:07Z | |
| dc.date.available | 2024-7-17T07:39:07Z | |
| dc.date.issued | 2024 | |
| dc.identifier.uri | http://ir.lib.ncu.edu.tw:444/thesis/view_etd.asp?URN=111423036 | |
| dc.contributor.department | 資訊管理學系 | zh_TW |
| DC.description | 國立中央大學 | zh_TW |
| DC.description | National Central University | en_US |
| dc.description.abstract | 為了促進與不同地區人們的交流，書信寫作變得普遍，隨著時代變遷，通訊方式變得更加多樣和便捷。從傳統信件、短信、電子郵件到消息應用程式的演變，文本使用的嚴謹程度各不相同，消息變得更加口語化。儘管這種轉變對人際交流並未構成重大問題，但在互聯網和深度學習應用的時代，文本內容的準確性和識別已成為關鍵問題。在依賴文本的應用程式如聊天機器人和搜索引擎中，拼寫錯誤會導致錯誤的判斷，妨礙預期結果的實現。因此，拼寫檢查具有至關重要的作用。以往的研究通常集中於簡體中文或繁體中文，這會因為這兩種書寫系統的差異而導致誤判。此外，俚語和縮寫的普及使當前模型無法解讀這些內容。為了解決這些挑戰，本研究介紹了一種基於擴散的新方法——DiffuCSC，旨在克服現有研究的局限性，提供更好的中文拼寫檢查和修正。 | zh_TW |
| dc.description.abstract | To facilitate communication with people in different locations, the prevalence of letter writing began, and as times changed, the modes of communication became more diverse and convenient. From traditional letters, text messages, and emails to the evolution of messaging apps, the rigor of text usage has varied, with messages becoming more conversational in nature. While this shift has not posed a significant problem for interpersonal communication, the accu-racy and recognition of text content have become critical issues in the internet era and deep learning applications. In text-reliant applications such as chatbots and search engines, incorrect spelling can lead to erroneous judgments, thwarting the intended outcomes. Hence, spelling checking holds paramount importance. Previous research often focused on either Simplified or Traditional Chinese, leading to misjudgments due to the differences between these scripts. Addi-tionally, the proliferation of slang and abbreviations presents content that is undecipherable by current models. To address these challenges, this study introduces a novel approach for mixed Chinese Spelling Checking based on diffusion—DiffuCSC, aimed at overcoming the limitations of existing research and providing improved Chinese spelling checking and correction. | en_US |
| DC.subject | 自然語言處理 | zh_TW |
| DC.subject | 中文拼寫檢查 | zh_TW |
| DC.subject | 擴散模型 | zh_TW |
| DC.subject | Natural Language Processing | en_US |
| DC.subject | Chinese Spelling Checking | en_US |
| DC.subject | Diffusion Model | en_US |
| DC.title | A Novel Diffusion-Based Spelling Checking on Hybrid Chinese Characters | en_US |
| dc.language.iso | en_US | en_US |
| DC.type | 博碩士論文 | zh_TW |
| DC.type | thesis | en_US |
| DC.publisher | National Central University | en_US |