What are the software requirement?

1. Python IDE

How to set up the python environment and install packages?

1. Clone code by clinking on the download button

2. Open the project folder with your choice of python IDE.
3. Execute below comment to install required python packages.

    pip install -r requirements.txt

Where to download the data ?

1. "GoogleNews-vectors-negative300.bin" online as word-embedding. https://drive.google.com/file/d/0B7XkCwpI5KDYNlNUTTlSS21pQmM/edit
2. Twitter API: https://developer.twitter.com/en

How to get the results?

1. Run "total_classification.py" with the csv data with the input column "tweet_text" and "category_id", where "tweet_text" is the tweet content will be trained and "category_id" is the category each tweets is labeled.

2. Line 93 "model.add(Dense(8))" is how many categories does the labeling have, change the input value accordingly.

3. Run the "total_classification.py" with the data will output a directory called "model_save", which will have three components of the model.

4. Run "model_prediction.py" to use the model saved in "model_save" directory, which will output the predication of tweeter content.

            a. Create a directory named 'data' to hold the input csv file.
            
            b. Create a directory named 'result' to save the output csv file.
   

Tutorial Video