ritikamotwani / Deception-Detection

Detecting Deception using Verbal Cues | Dataset Used: Real life trial data collected during a series of experiments at Michigan (http://web.eecs.umich.edu/~zmohamed/PDFs/Trial.ICMI.pdf) and Deceptive Opinion Spam Corpus v1.4(https://myleott.com/op-spam.html)

Deception-Detection

Classifiers added to detect deception through transcripts.
Dataset Used:

1. Real life trial data collected during a series of experiments at Michigan (http://web.eecs.umich.edu/~zmohamed/PDFs/Trial.ICMI.pdf) (Folder - dataset)
2. Deceptive Opinion Spam Corpus v1.4 (https://myleott.com/op-spam.html)

Installation

Run the following command to install all python packages that'll be used in the project:

pip install -r requirements.txt

Approaches

Opinion Spam Dataset:

Folder: OpinionSpam

1. Parameters: NGram Approach Classifiers: SVM, NB, Random Forest

   To execute:
      python3 Ngrams_And_Classifiers.py 

SVM


NB


Random Forest

2. Parameters: LIWC Classifiers: SVM, NB, Random Forest

   To execute:
      python3 LIWC_And_Classifiers.py 

NB


SVM


Random Forest

3. Parameters: NGrams, LIWC Classifiers: SVM

   To execute:
      python3 SVM_Ngrams_LIWC.py

4. Recurrent Neural Networks

   To execute:
      python3 RNN

Real Life Dataset

Folder: RealLife

1. Classifiers: SVM, NB

   To execute:
      python3  Classifiers.py

SVM


NB

2. Parameters: NGram Approach Classifiers: SVM, NB, Random Forest

   To execute:
      python3 Ngrams_And_Classifiers.py 

SVM


NB


Random Forest

3. Parameters: LIWC Classifiers: SVM, NB, Random Forest

   To execute:
      python3 LIWC_And_Classifiers.py 

NB


SVM


Random Forest

4. RNN

   To execute:
      python3 RNN.py