senti_c (sentiment analysis toolkit for traditional Chinese)
簡介
本工具為繁體中文情感分析套件,支援三種類型分析:句子情感分類、屬性術語提取、屬性情感分類;同時提供函數供使用者應用其它資料重新微調模型。
目錄
執行環境
- python3.8
安裝方式
- 要先至 https://pytorch.org/ 下載適合作業系統的PyTorch Version 1.x套件。
1.pip
pip install senti_c
2.from source
git clone https://github.com/hsinmin/senti_c
cd senti_c
python3 setup.py install
功能介紹
1.句子情感分類:預測
from senti_c import SentenceSentimentClassification
sentence_classifier = SentenceSentimentClassification()
test_data = ["我很喜歡這家店!超級無敵棒!","這個服務生很不親切..."]
result = sentence_classifier.predict(test_data,run_split=True,aggregate_strategy=False) # 可依據需求調整參數
- 結果如下:
2.句子情感分類:重新微調模型
from senti_c import SentenceSentimentModel
sentence_classifier = SentenceSentimentModel()
sentence_classifier.train(data_dir="./data/sentence_data",output_dir="test_fine_tuning_sent") # 可依據需求調整參數
3.屬性情感分析:預測
from senti_c import AspectSentimentAnalysis
aspect_classifier = AspectSentimentAnalysis()
test_data = ["我很喜歡這家店!超級無敵棒!","這個服務生很不親切..."]
result = aspect_classifier.predict(test_data,output_result="all") # 可依據需求調整參數
- 結果如下:
4.屬性情感分析:重新微調模型
from senti_c import AspectSentimentModel
aspect_classifier = AspectSentimentModel()
aspect_classifier.train(data_dir="./data/aspect_data",output_dir="test_fine_tuning_aspect") # 可依據需求調整參數
範例程式
相關功能demo可參考examples資料夾中的function_demo檔案。
資料
本研究蒐集Google評論上餐廳與飯店領域評論內容、並進行句子情感分類、屬性情感分析標記 (屬性標記與情感標記)。
相關資料格式請見data資料夾。
引用
1.論文:
凃育婷(2020)。基於順序遷移學習開發繁體中文情感分析工具。國立臺灣大學資訊管理學研究所碩士論文,台北市。
2.實驗室:
Business Analytics and Economic Impact Research Lab
Department of Information Management
National Taiwan University
http://www.im.ntu.edu.tw/~lu/index.htm
致謝
本套件基於 Hugging Face 團隊開源的 transformers。
Metadata
Release files for senti-c 0.1.2
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| senti_c-0.1.2.tar.gz | 40.8 kB | Without provenance | Details for senti_c-0.1.2.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| senti_c-0.1.2-py3-none-any.whl | Python 3 | none | any | Details for senti_c-0.1.2-py3-none-any.whl |
Total release size: 89.7 kB
Release files / senti_c-0.1.2.tar.gz
| Download URL | senti_c-0.1.2.tar.gz |
|---|---|
| Size | 40.8 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
90c8bd7be157d458cbfda44c7cd57920e8e2a1d592fb10f731ff56e42408d25d
|
|
BLAKE2b-256 checksum How to use checksums |
c7223c0c11500672518445e9cdc88e292613ec009720bd1b342163bfe016c5c6
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/3.4.1 importlib_metadata/3.10.0 pkginfo/1.7.0 requests/2.22.0 requests-toolbelt/0.9.1 tqdm/4.48.2 CPython/3.8.5
|
Release files / senti_c-0.1.2-py3-none-any.whl
| Download URL | senti_c-0.1.2-py3-none-any.whl |
|---|---|
| Size | 48.9 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
6ac0031f1f764c80a34b26e79546fc18bb41e69bdce3288a688a21380fd885fa
|
|
BLAKE2b-256 checksum How to use checksums |
35c05d189b78ec82eadbd357a67db74800396327f4cd48512f349908bd956a08
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/3.4.1 importlib_metadata/3.10.0 pkginfo/1.7.0 requests/2.22.0 requests-toolbelt/0.9.1 tqdm/4.48.2 CPython/3.8.5
|
Security for senti-c 0.1.2
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.