Experiment with tokenisation, simple sentiment rules, and text classification sandboxes. Learn how AI systems interpret and structure language.
Tokenization Playground
Paste any English text below. This tool will:
• Lowercase the text
• Strip basic punctuation
• Split into whitespace-delimited tokens
• Compute token count and simple frequency statistics
Tokenised Output:
Tokens: — | Unique: —
Sentiment Analyzer (Rule-Based)
This is a simple lexical sentiment model: it counts positive and negative words using a handcrafted dictionary, then labels the text as Positive, Negative or Neutral.
Overall Sentiment: —
Negative score: 0
Net score: 0
Matched Words:
Simple Text Classifier
This rule-based classifier tries to guess a topic label (Tech, Health, Finance) using keyword matching. This simulates the idea behind text categorisation.
Predicted Category:
Word Similarity Sandbox
This demo assigns random 3-dimensional vectors to words and computes a cosine similarity score (simulating how embeddings work conceptually).