README.md

metadata

title: ExcelWordCloud
emoji: 🐨
colorFrom: blue
colorTo: gray
sdk: gradio
sdk_version: 4.21.0
app_file: app.py
pinned: false
license: mit

This Gradio app enables users to generate sentiment-based word clouds from text data in Excel files. It visually represents positive and negative sentiments using color-coded words, providing an intuitive analysis of textual data. Here's a brief guide on how to use it:

How to Use

1. Upload Excel File: Click the "Upload Excel File" button to select and upload your Excel file. The file should contain textual data you wish to analyze.
2. Enter Column Name: Input the name of the column from your Excel file that contains the text data. This tells the app which column to use for generating the word cloud.
3. Generate Word Cloud: After uploading the file and specifying the column name, submit the form. The app will process the text data, identifying positive and negative words to generate a word cloud. Positive words are displayed in green, negative words in red, and neutral or unspecified words in gray.
4. View Results: The generated word cloud will be displayed as the output. This visual representation helps you quickly grasp the overall sentiment of the text data.

Technical Details

• The app uses the opinion_lexicon from NLTK to distinguish between positive and negative words.
• A custom color function, SimpleGroupedColorFunc, assigns specific colors to words based on their sentiment.
• The wordcloud library generates the visual representation, which is then recolored according to the sentiment analysis.
• Gradio's interface (gr.Interface) creates a simple web app for interacting with the Python function, allowing for easy file upload and parameter input.

Notes

• Make sure the column name you enter matches exactly with one in the Excel file, including case sensitivity.
• The word cloud provides a snapshot of the sentiment but does not convey the context or the intensity of sentiments.

This tool is particularly useful for quickly analyzing customer feedback, reviews, or any textual data where understanding sentiment is valuable.