ChordReviewsVis 0.2.9

Process reviews data, apply text preprocessing, and generate a chord plot visualization showing word co-occurrence patterns and sentiment analysis.

Chord Reviews

Overview

ChordReviewsVis is a Python package designed to process and visualize review data by generating chord plots. These visualizations illustrate word co-occurrence patterns and sentiment analysis, providing insights into the textual data. For this, it relies on the following features:

• Labels for each node: The top nouns and adjectives extracted from the reviews were displayed around the graphic.
• Label bars: Below the labels, there is a bar whose color illustrates the word’s overall frequency in reviews. The darker the color, the more frequent the word.
• Edges: The line connecting the words that occur together.
• Edge thickness: This characteristic shows how often the connected words appear in the same sentence. The more often they are together, the thicker the line is.
• Edge color: The color shows the overall sentiment of the words that are being connected. Red was used for negative sentiments, blue for neutral ones, and green for positive sentiments.

This package was developed by Felix Jose Funes as part of his master's dissertation at Universidade Nova de Lisboa, which was supervised by Prof. Nuno Antonio, PhD.

Installation

To install ChordReviewsVis, use pip:

pip install ChordReviewsVis

Usage

First, import the necessary libraries and the ChordReviews function:

import pandas as pd
from ChordReviewsVis import ChordReviews

Prepare the DataFrame with a text column containing review data. Then call the ChordReviews function:

# Load DataFrame
df = pd.read_csv("filepath")

# Generate chord plot
ChordReviews(df, 'review')

Some datasets that can be used for this purpose are:

• IMDB Movie Reviews
• Women's E-Commerce Clothing Reviews
• Amazon Fine Food Reviews

Function Parameters

• df (pandas.DataFrame): DataFrame containing review data.
• text_column (str): Name of the column containing the text data.
• size (int, optional): Size of the output chord plot. Default is 300.
• stopwords_to_add (list, optional): Additional stopwords to include in the stop words set. Default is an empty list.
• stemming (bool, optional): Whether to apply stemming to words. Default is False.
• lemmatization (bool, optional): Whether to apply lemmatization to words. Default is True.
• words_to_replace (dict, optional): A dictionary where keys are words to be replaced and values are the replacements. Default is an empty dictionary.
• label_text_font_size (int, optional): Font size for the labels in the chord plot. Default is 12.

Returns

• hv.Chord: A chord plot visualization of word co-occurrence patterns and sentiment analysis.

Examples

Basic Usage

# Import necessary libraries
import pandas as pd
from ChordReviewsVis import ChordReviews

# Load dataset
df = pd.read_csv("https://github.com/felix-funes/ChordReviewsVis/raw/main/Test%20Dataset%20-%20IMDB%20Movie%20Reviews.csv")

# Generate chord plot
ChordReviews(df, 'review')

Custom Parameters

Though lemmatization is used by default, users have the possibility of using stemming if it suits their needs better.

# Import necessary libraries
import pandas as pd
from ChordReviewsVis import ChordReviews

# Load dataset
df = pd.read_csv("https://github.com/felix-funes/ChordReviewsVis/raw/main/Test%20Dataset%20-%20IMDB%20Movie%20Reviews.csv")

# Generate chord plot
ChordReviews(df, 'review', stemming=True, lemmatization=False)

To refine the visualization, it is possible to use the "stopwords_to_add" parameter to remove irrelevant words and "words_to_replace" to unify terms with the same meaning.

# Import necessary libraries
import pandas as pd
from ChordReviewsVis import ChordReviews

# Load dataset
df = pd.read_csv("https://github.com/felix-funes/ChordReviewsVis/raw/main/Test%20Dataset%20-%20IMDB%20Movie%20Reviews.csv")

# Generate chord plot
chord_reviews(df, 'Review', stemming=False, lemmatization=True, stopwords_to_add=["wa", "ha"], words_to_replace={"movie": "film"})

Because of the prevalence of the words "film" and "movie", they may be considered stop words. It is possible to remove them using the parameter "stopwords_to_add". For presentation purposes, the final plot and label text can be resized.

# Import necessary libraries
import pandas as pd
from ChordReviewsVis import ChordReviews

# Load dataset
df = pd.read_csv("https://github.com/felix-funes/ChordReviewsVis/raw/main/Test%20Dataset%20-%20IMDB%20Movie%20Reviews.csv")

# Generate chord plot
chord_reviews(df, 'Review', stemming=False, lemmatization=True, stopwords_to_add=["wa", "ha", "movie", "film"], label_text_font_size=13, size=400)

Dependencies

Ensure you have the following libraries installed:

• pandas
• numpy
• nltk
• BeautifulSoup
• re
• holoviews

These can be installed via pip:

pip install pandas numpy nltk beautifulsoup4 re holoviews

License

This project is licensed under the MIT License.

Contact

For any issues or inquiries, please contact the package maintainer via LinkedIn.

---

By using this package, you agree to the terms outlined in the LICENSE file included in the repository.