Autocomplete

• A simple autocompletion system that uses the Brown University corpus to generate suggestions.

Install Dependencies

• Install Python 3.7 (add to PATH)
• Run pip3 install -r requirements.txt
• Run py -3 download_nltk_packages.py

Setup

1. Run py -3 create_unigrams.py to process the Brown University corpus.
2. Run py -3 important_unigrams.py add the most important unigrams to the suggestions.
3. Run py -3 important_bigrams.py then py -3 important_trigrams.py to do the same for bigrams and trigrams.
4. Run py -3 create_trie.py to create the trie of suggestions.
5. Run py -3 predict_text.py to open a tkinter window to try it out.

How to Use

Type in something and press tab to use a suggestion. You might need to type something specific to the Brown University corpus since the variety of documents is limited.

Example

Directories

• /data Where the processed corpus data is saved
• /graphs Images of the graphs created by important_unigrams.py, important_bigrams.py, and important_trigrams.py