Real-time writing prediction and correction model based on the trie data structure, the Boyer-Moore algorithm for string matching, and the Levenshtein distance (using dynamic programming) for string similarity.
This was designed and implemented as the final project of the "Design and Analysis of Algorithms" and "Programming for Data Analysis" classes which are part of the Master of Computer Systems in the Western Institute of Technology and Higher Education ITESO University.
Clone this repository in your preferred path.
git clone https://github.com/EarlOrlando/WordsSentencesPrediction.gitInstall all Python dependencies. This project may work with other versions but this project has been tested only with the shown below.
python -m pip install requests==2.25.0
python -m pip install pandas==1.1.5
python -m pip install numpy==1.19.3
python -m pip install bs4==4.9.3
python -m pip install nltk==3.5
python -m pip install matplotlib==3.3.3By executing all the Jupyter notebook cells the example GUI is run.
File sentences.csv stores all the pages acquired from Wikipedia en Español so these are not downloaded on every execution. This file has to be removed in order to download new Wikipedia examples.
File Crea_total.zip contains the Spanish language words corpus acquired from Corpus de Referencia del Español Actual (CREA).