A fast Chinese event extraction program 一個快速實現事件抽取的程式，繁體中文與簡體中文都支持
- 若未分詞，用jieba分詞(速度快是優勢)
- 運用pylyp進行詞性標註跟依存句法分析
- 建立找中心語(Head)規則
- 建立找前後論元規則
- 建立事件抽取擴充版規則
-
opencc : a tool for converting traditional Chinese charaters into simplified Chinese charaters. https://github.com/yichen0831/opencc-python
-
jieba : an efficient Chinese tokenization tool. https://github.com/fxsjy/jieba
-
pyltp : a traditional Chinese language preprocessing utility containing tokenization, p.o.s. tagging, ne recognition , dependency parsing, etc. tools. https://github.com/HIT-SCIR/pyltp
After Installation, you should have a ltp_data folder containing parser.model and pos.model ready for loading
>>> from Event_Extraction import Event_Extraction
1.沒有分詞
>>> events = Event_Extraction('三商美邦总经理杨棋材，请辞获准。', tra_sim = True, tokenize = False, expand = True)
2.有分詞(以空白間隔)
>>> events = Event_Extraction('三商 美邦 总经理 杨棋材 ，请辞 获准 。', tra_sim = True, tokenize = True, expand = True)
--------
3.繁體中文輸入
>>> events = Event_Extraction('三商 美邦 總經理 楊棋材 ，请辭 獲准 。', tra_sim = False, tokenize = True, expand = False)
>>> events.events
1. standard version:
[['总经理','杨棋材'], ['请辞','获准']]
2. expand version:
[['美邦','总经理','杨棋材'], ['请辞','获准']]