https://github.com/marxmit7/twitanalysis

Twit sentiment analysis using deep learning
https://github.com/marxmit7/twitanalysis

analysis deep keras learning nlp python tensorflow twitter

Last synced: about 7 hours ago
JSON representation

Twit sentiment analysis using deep learning

• Host: GitHub
• URL: https://github.com/marxmit7/twitanalysis
• Owner: marxmit7
• Created: 2018-07-31T20:59:36.000Z (over 6 years ago)
• Default Branch: master
• Last Pushed: 2018-08-01T07:27:02.000Z (over 6 years ago)
• Last Synced: 2024-08-16T13:02:43.184Z (3 months ago)
• Topics: analysis, deep, keras, learning, nlp, python, tensorflow, twitter
• Language: Jupyter Notebook
• Homepage:
• Size: 10.6 MB
• Stars: 2
• Watchers: 0
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: readme.md

Awesome Lists containing this project

README

        
# Twitter sentiment analysis

### Dataset used

[Thinkbook](http://thinknook.com/wp-content/uploads/2012/09/Sentiment-Analysis-Dataset.zip)

### How to Use

```

pip3 install -r requirements.txt

python useit.py

```

## Self.blog

I used this network architecture

![model.png](model.png)

On 5th epoch , validation accuracy came out as 79.36 %

It took me around ~ 3 hour to train the model .

This project is quite simple and interesting. Here are some examples that I tested.

Input a twit to be evaluated  :  `My deep learning model is working so fine`

`positive sentiment 91.65% confidence `

Input a twit to be evaluated  :  `I love fashion`

`positive sentiment 91.50% confidence`

Input a twit to be evaluated  :  `i don't want this to  happen again`

`negative sentiment 90.93% confidence `

### Developer

    

        Amit Kumar











Want to contribute?

Just raise an issue and Give PR !