All projects
Data Analysis & Visualization

Word Cloud Generator

A tool that turns text data into a visual summary

localhost
Word Cloud Generator
Streamlit
Interface
NLP
Domain
2023
Year

Overview

It is hard to tell what a pile of text is about at first glance. With open-ended survey answers, customer reviews or social media posts, the first need is not detailed analysis but a quick overview.

The application takes the text and cleans it first: punctuation, numbers and meaning-free connective words (stop words) are stripped. Word frequencies are then computed and a word cloud is generated at sizes proportional to those frequencies. The user can change the colour palette, maximum word count and shape mask to make the output presentation-ready.

Because it is built with Streamlit, users can run it in the browser with no installation. The tool provides a practical first step for seeing which themes dominate before moving into deeper text analysis.

Highlights

  • Automatic text cleaning and stop word removal
  • Word clouds sized by frequency
  • Colour palette, word count and shape mask options
  • A browser interface requiring no installation

Objectives

  • Quickly surface the dominant themes in a body of text
  • Produce presentation-ready visual output
  • Open a door to text analysis for non-technical users

Solution

  • A preprocessing pipeline: cleaning, normalisation, stop words
  • Frequency computation and weighting
  • A parametric visualization engine
  • Instant preview with Streamlit

Approach

  1. 01Defining preprocessing rules
  2. 02Designing the visualization parameters
  3. 03Testing across different text types

Outcomes

  • A body of text became interpretable in seconds
  • Presentation-ready visuals were produced
  • A fast discovery step before deeper analysis

Challenges

  • The adequacy of the Turkish stop word list
  • Merging word stems
  • Keeping visual density from hurting readability

Selected Work

Similar projects

View all