Skip to content

polabroda/NLP_manipulation_detection

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

12 Commits
 
 
 
 

Repository files navigation

NLP Manipulation Detection

This project focuses on detecting psychological manipulation patterns in conversations using Natural Language Processing and Machine Learning techniques.

The solution includes text preprocessing, vectorization, exploratory analysis, model training, and classification performance evaluation for manipulation-related language patterns.


Project Overview

  • Natural Language Processing workflow implementation
  • Text preprocessing and cleaning
  • TF-IDF text vectorization
  • Manipulative language classification
  • Exploratory text analysis
  • Model training and evaluation
  • Class distribution analysis
  • Performance metric comparison

NLP Workflow

The project includes the following stages:

Data Preprocessing

The preprocessing pipeline includes:

  • Text cleaning
  • Lowercasing
  • Missing value handling
  • Removal of unnecessary characters
  • Text normalization

Feature Engineering

Text data was transformed using:

  • TF-IDF vectorization
  • Numerical text representation
  • Sparse feature matrices for machine learning models

Machine Learning

The project focuses on training classification models capable of identifying manipulation patterns in conversations.

The workflow includes:

  • Model training
  • Prediction generation
  • Classification evaluation
  • Model comparison

Dataset

The project uses the following Kaggle dataset:

Psychological Manipulation Conversations Dataset

https://www.kaggle.com/datasets/tatheerabbas/psychological-manipulation-conversations-dataset

The dataset is not included in this repository because of its size.


Analysis Scope

The project analyzes:

  • Manipulative language patterns
  • Class distribution
  • Text frequency characteristics
  • Linguistic patterns in conversations
  • Classification effectiveness for manipulation detection

Technologies

  • Python
  • Pandas
  • NumPy
  • Scikit-learn
  • Natural Language Processing
  • TF-IDF Vectorization
  • Matplotlib
  • Seaborn
  • Jupyter Notebook

Goal

The goal of this project is to demonstrate practical Natural Language Processing and Machine Learning skills for detecting manipulative communication patterns in textual data.


Results

The solution successfully demonstrates:

  • End-to-end NLP workflow implementation
  • Text preprocessing and feature engineering
  • TF-IDF vectorization
  • Classification model training
  • Manipulation pattern analysis
  • Machine learning evaluation workflows
  • Data exploration and visualization
  • Practical NLP pipeline development

Author

Paulina Broda

About

NLP project focused on detecting manipulation in conversations using machine learning techniques.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors