It's a Chrome Extension, which can translate Chinese words into Taiwanese words.
By analyzing web content, the extension will translate Chinese words into Taiwanese words, and highlight which words were translated, and show what the original words and english words are.
It allows users to open menu to add a feedback by a right-click.
這是一個能夠翻譯並標示出大陸用語的 Chrome 瀏覽器擴充套件
透過分析網頁內容,將大陸用語直接換成台灣用語,並在頁面中標示出哪些是已經被置換過,且顯示原本使用的用語和英文用語是為何。
允許使用者透過點擊右鍵開啟選單來新增回饋。
It will release on Extension Store You can find it on Extension Store
git clone https://github.com/NCNU-OpenSource/Shape_of_Taiwan.git
- 開啟 "開發人員模式"
- 載入未封裝的擴充功能,選擇專案資料夾
- 預設擴充功能為開啟狀態,若要關閉功能,點擊擴充功能的 icon 即可
- 安裝專案所需的軟體和套件
- 執行
backend
資料夾中的shape_of_taiwan_backend.py
- 將
content.js
中ajax
的url
改成本機的url
- Chrome Extensions Framework
- JavaScript
- JSON
- Regular Expression
- Python Flask
- uwsgi
- certbot
- nginx proxy pass
- XMLHttpRequest
- Cross Origin Resource Sharing
- Jieba
- gensim
{
"內存": {
"zh_TW": "記憶體",
"en_US": "RAM"
},
"報文": {
"zh_TW": "封包",
"en_US": "Packet"
},
"光驅": {
"zh_TW": "光碟機",
"en_US": ""
}
}
The backend folder should not put in production It's here only for development
- 使用 LSI 計算文章間(網頁 & 主題文章)的相似度,取 0.7 作為標準判斷文章主題
- 資料預處理與訓練範例檔案
backend/static/training_tool/data_prepare.py {訓練資料檔名}
- 產出 cleaned_training_data.datset
- 為去除無用符號、替換同義字、斷詞後 dataset
backend/static/training_tool/train_prepare.py
- 產出用於擴充套建 .dic(dictionary) .lsi(lsi model) .mm(corpus) 的模型
- 功能預期成果 & 模型調整相關說明共筆