Create a python environment with conda create -n mmdet python=3.6
.
Install python libraries with pip install -r requirements.txt
.
We get data from FakeNewsNet, and format it to a .csv
file, with the data_process/generate_csv.py
.
Details are written in the comments in the python file.
Sentiment analysis is done by data_process/sentiment_analysis.py
, and is visualized by data_process/sentiment_plot.py
. data_process/tsne.py
is used to visualize tsne result.
Models are implemented in the infer/infer.py
.
Use infer/grid_tuning.py
to search for hyper parameters.
Images are in img/.
.
Report is available at JBox and OneDrive.
Here comes the ppt.