PyLogicBot 0.2

A module for analyzing presidential speeches

PyChatBot Presidential Speech Analysis EFREI

Overview

This Python script analyzes the speeches of various presidents. It performs pre-processing on the speech texts and provides basic functions for text analysis, including the calculation of TF-IDF (Term Frequency-Inverse Document Frequency) matrices.

Project Structure

The project is structured as follows:

• app.py : The main application file.
• Basic_Functions.py : Contains basic functions used in the application.
• Cosine_Similarity_Functions.py : Contains functions for calculating cosine similarity.
• Features_to_be_Developed.py: Contains functions for features to be developed.
• Menu.py: Contains the application menu.
• Pre_Processing.py : Contains functions for data pre-processing.
• tokenization_question.py : Contains functions for tokenizing questions.
• cleaned/ : Contains cleaned presidential speeches.
• speeches/ : Contains original presidential speeches.
• static/ and templates/: Contain files for the application's user interface.

Pre-Processing

1. Extract President Names: Extracts the names of the presidents from the filenames of speech texts.
2. Associate First Name: Associates a first name with each president.
3. Display President Names: Displays the list of president's names, avoiding duplicates.
4. Text Conversion: Converts the original speech texts to lowercase and creates cleaned versions in the "cleaned" folder.
5. Remove Punctuation: Removes punctuation characters from the cleaned speech texts.

Basic Functions

• Proportion in Documents: Calculates the proportion of documents containing a given word.
• Get Unique Words: Retrieves a list of unique words from the cleaned texts.
• Get TF Matrix: Generates a Term Frequency (TF) matrix for the words in the speeches.
• Get IDF Scores: Calculates the Inverse Document Frequency (IDF) scores for the unique words.
• Get TF-IDF Matrix: Calculates the TF-IDF matrix using the TF matrix and IDF scores.
• Transpose Matrix: Transposes a given matrix.

Features

1. Display Least Important Words

• Function: get_least_important_words(directory)
• Description: Retrieves words with TF-IDF score equal to 0.

2. Display Highest TF-IDF Score

• Function: display_highest_tfidf_score(directory)
• Description: Identifies words with the highest TF-IDF scores.

3. Most Repeated Words

• Function: most_repeated_words(directory)
• Description: Identifies the most repeated words in the speeches.

4. First President to Talk About a Word

• Function: first_president_word(word)
• Description: Identifies the first president to mention a specific word.

5. President who said the most a word

• Function: president_who_said_it_the_most(word,directory)
• Description: Identifies the president who said a word the most.

6. Get president names

• Function: extract_presidents_names(directory)
• Description: Display the names of the presidents.

Usage

1. Clone this repository.
2. Install dependencies with pip install <name of the dependency>.
3. Run python app.py or flask run to start the application.
4. Open your browser at http://localhost:5000 to use the application.

Menu

• The script includes a menu-driven interface for user interaction. Users can choose specific tasks from the menu.

Dependencies

• Python 3.x >= 3.9
• Flask
• Random
• Math
• OS
• Re

Authors

Andrea CHARVIERE and Antonin LARTILLOT-AUTEUIL