Title: Sentiment Analyzer 😁🙂😒😑
Sketch link:
One sentence description:
🧨 A tool to help you analyze text sentiment to help you speak with others using words they connect to.
Project summary:
🧨 This project is a Sentiment Analyzer built on the theoretical principle of Mutual Information, which measures the relationship between data points. The tool explores how individual words in a piece of text contribute to overall sentiment. By analyzing text word-by-word, the system evaluates whether each word is positive, negative, or neutral and uncovers patterns in word usage linked to emotional states.
Users can input any text into the tool, such as a message, journal entry, or comment, and click "Analyze Sentiment." The program processes the input and breaks it into words, using a sentiment analysis algorithm to assign a score to each one. Words associated with positive emotions (e.g., "happy," "love") are highlighted in blue, negative words (e.g., "hate," "angry") in red, and neutral words (e.g., "and," "is") in gray.
As a unique feature, the project tracks word usage based on sentiment over time. For example, if someone often says “yippie” when they are happy, the tool identifies this trend and categorizes “yippie” as a positive word. By keeping a record of words for each sentiment, the system helps users gain deeper insights into how they or others express emotions in language.
The project demonstrates the concept of mutual information by revealing the relationship between specific words and emotional states. Words that appear consistently in certain emotional contexts hold higher informational value for that sentiment.
This tool is practical for self-reflection, communication studies, and even understanding patterns in social interactions. Its clean interface and dynamic visualization make it user-friendly and engaging for anyone interested in exploring language and emotion.
Inspiration: How did you become interested in this idea? Quotes, photographs, products, projects, people, music, political events, social ills, etc:
🧨 My inspiration came from an unlikely place. Professor Shiffman shared a video with me about cosine similarity, as we discussed about it in class. One thing led to another, and I stumbled upon a video discussing Mutual Information. I loved it. And the first idea that came to my mind was this.
Process: How did you make this? What did you struggle with? What were you able to implement easily and what was difficult?
🧨 This project started with my final project proposal:
[https://docs.google.com/presentation/d/1LVLoCoD9h8W47YOI1d-uVYpUI8tzWUPvxMKnxr5VMxc/edit#slide=id.p](https://docs.google.com/presentation/d/1LVLoCoD9h8W47YOI1d-uVYpUI8tzWUPvxMKnxr5VMxc/preview#slide=id.p)
Initially, I had three ideas I wanted to play with. But with feedback from class, and overall sentiment (no pun intended), I decided to go with the sentiment analyzer idea. It was practical and fun.
🧨 Prior to the play test, I started this project in the realms of VS Code.
and while I had high hopes for it to work on the first try, I DID NOT!!
Everything seemed to go well, but for some reason, the “Analyze Sentiment” button did not work. This was frustrating.