This project applies LLM models and 2 traditional machine learning models to perform fact-checking prediction. It explores different ML algorithms to optimize the evidence retrievals and fact-checking classifications.
- Data Preprocessing 1
- Fine Tuning
- Evidence Prediction
- Evidence Prediction Training
- Predictions
- Class Predictions
- Data Preprocessing 2
- SVM Prediction
- LSTM Prediction
- Output
Describe the first data preprocessing steps, including data cleaning, normalization, etc.
Detail the fine-tuning processes, including model selection, parameter adjustments, etc.
Explain the methodology and objectives of the evidence prediction phase.
Discuss the training process for evidence prediction, including data used, training techniques, etc.
Describe how predictions are made, the algorithms used, and any relevant details.
Elaborate on the classification approach, including the classes predicted and the methods employed.
Detail additional data preprocessing steps, if different from the first phase.
Describe the use of Support Vector Machine (SVM) for predictions, including model configuration and performance.
Explain the implementation of Long Short-Term Memory (LSTM) networks, focusing on its setup, training, and application.
Discuss the output of the project, including how results are presented and interpreted.