Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/jung217/lstm_bert_sentiment_analysis
基於LSTM&BERT機器學習之網路輿情分析
https://github.com/jung217/lstm_bert_sentiment_analysis
ai analysis bert lstm machine-learning network shopee tensorflow
Last synced: 12 days ago
JSON representation
基於LSTM&BERT機器學習之網路輿情分析
- Host: GitHub
- URL: https://github.com/jung217/lstm_bert_sentiment_analysis
- Owner: Jung217
- License: mit
- Created: 2023-05-31T08:49:20.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2023-05-31T12:32:02.000Z (over 1 year ago)
- Last Synced: 2024-11-25T01:12:59.321Z (2 months ago)
- Topics: ai, analysis, bert, lstm, machine-learning, network, shopee, tensorflow
- Language: Jupyter Notebook
- Homepage:
- Size: 56.4 MB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
## Network Sentiment Analysis Based on LSTM & BERT Machine Learning
### 基於 LSTM & BERT 機器學習之網路輿情分析
> 「縱浪大化中，不喜亦不懼。應盡便須盡，無復獨多慮。」－陶淵明〈神釋〉
> 「AI科技發展快速，其無非是世界一道不可阻攔的洪流，人們應保持開放樂見的心態面對，在一波波的浪潮中，尋等機會，一舉站上AI的浪頭上，盡享AI帶來的便利及紅利。」－CCJ
## Preface
本專案為紀念2023年5月，整個月不捨晝夜訓練模型的我，跟訓練AI模型差點成為Colab Pro的我，僅此。
## Introduction
* [Detail Report](https://github.com/Jung217/LSTM_BERT_Sentiment_Analysis/blob/main/Network%20Sentiment%20Analysis%20Based%20on%20LSTM%20%26%20BERT%20Machine%20Learning_By_CCJ.pdf) : All the details about this project
* [LSTM Model](https://github.com/Jung217/LSTM_BERT_Sentiment_Analysis/tree/main/LSTM)
* [LSTM.ipynb](https://github.com/Jung217/LSTM_BERT_Sentiment_Analysis/blob/main/LSTM/LSTM.ipynb) : Main training and testing program
* [input](https://github.com/Jung217/LSTM_BERT_Sentiment_Analysis/tree/main/LSTM/input)
* [apply-jieba-tokenizer](https://github.com/Jung217/LSTM_BERT_Sentiment_Analysis/tree/main/LSTM/input/apply-jieba-tokenizer) : Tokenizer and testing data
* [fake-news-pair-classification-challenge](https://github.com/Jung217/LSTM_BERT_Sentiment_Analysis/tree/main/LSTM/input/fake-news-pair-classification-challenge) : Origin training data
* [BERT Model](https://github.com/Jung217/LSTM_BERT_Sentiment_Analysis/tree/main/BERT)
* [BERT.ipynb](https://github.com/Jung217/LSTM_BERT_Sentiment_Analysis/blob/main/BERT/BERT.ipynb) : Main training and testing program
* [commentss.csv](https://github.com/Jung217/LSTM_BERT_Sentiment_Analysis/blob/main/BERT/commentss.csv) : Training, Testing and validation data
* [Shopee Crawler(蝦皮爬蟲程式)](https://github.com/Jung217/LSTM_BERT_Sentiment_Analysis/tree/main/Shopee%20Crawler)
* [shopee.ipynb](https://github.com/Jung217/LSTM_BERT_Sentiment_Analysis/blob/main/Shopee%20Crawler/shopee.ipynb) : Crawler program to get data from [Shopee](https://shopee.tw/)
* [Data](https://github.com/Jung217/LSTM_BERT_Sentiment_Analysis/tree/main/Shopee%20Crawler/Data) : Commodity data from shopee
## Reference
1. [Kaggle-WSDM - Fake News Classification](https://www.kaggle.com/competitions/fake-news-pair-classification-challenge)
2. [LeeMeng:進擊的 BERT：NLP 界的巨人之力與遷移學習](https://leemeng.tw/attack_on_bert_transfer_learning_in_nlp.html)
3. [Evaluation Metrics : 分類模型](https://medium.com/ai%E5%8F%8D%E6%96%97%E5%9F%8E/evaluation-metrics-%E5%88%86%E9%A1%9E%E6%A8%A1%E5%9E%8B-ba17ad826599)
4. [「蝦皮爬蟲」｜商品資料＋留言評論](https://marketingliveincode.com/classification/crawler_king/110)