LYULU2 / FakeNewsDetection

Compared accuracies using FNDNet and LSTM neural networks. Compared performances using ELMo and GloVe embedding.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Fake news dectection with different embeddings and neural networks

create_dataset.ipynb is the code for building my datasets.
baseline.ipynb is the code for building the initial baseline model.
chi_sqaured_test.ipynb is the code for selecting the features.
improved_efficiency_baseline.ipynb is the code for baseline model with feature selection.
data_exploratory_analysis.ipynb is the code for initial data analysis.
Elmo_and_FNDNet.ipynb is the code for building architecture of ElMo and FNDNet.
Elmo_and_LSTM.ipynb is the code for building architecture of ElMo and LSTM.
Glove_and_FNDNet.ipynb is the code for building architecture of GloVe and FNDNet.
Glove_and_LSTM.ipynb is the code for building architecture of GloVe and LSTM.
grammar structure depth.ipynb is the code for analyzing the grammatical dependencies.
remove_high_chi2_score_word_test.ipynb is the code for further data analysis and experiment.
topic_modeling.ipynb is the code for topic modelling.

The implementation codes are included in the files described above. But some actual experiments results might appear random for the convenience of code-excecuting purpose.

About

Compared accuracies using FNDNet and LSTM neural networks. Compared performances using ELMo and GloVe embedding.


Languages

Language:Jupyter Notebook 100.0%