https://github.com/mgla96/authorpredictor
Predicts Anonymous Authors with Multinomial Naive Bayes Multi Classification
https://github.com/mgla96/authorpredictor
Last synced: 3 months ago
JSON representation
Predicts Anonymous Authors with Multinomial Naive Bayes Multi Classification
- Host: GitHub
- URL: https://github.com/mgla96/authorpredictor
- Owner: Mgla96
- Created: 2020-06-13T20:52:46.000Z (about 6 years ago)
- Default Branch: master
- Last Pushed: 2020-10-15T17:01:36.000Z (almost 6 years ago)
- Last Synced: 2025-12-26T18:48:27.162Z (7 months ago)
- Language: Python
- Homepage:
- Size: 4.36 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Author Predictor
This is a supervised learning author predictor program which I built to gain more knowledge in machine learning. This program utilizes Multinomial Naive Bayes Multi Classification with a bag-of-words representation of the text. I implemented bigrams and trigrams as well as used the nltk library to incorporate lemmetization as methods to further improve my accuracy. The accuracy could be further improved by implementing other features such as article length, using stop words, adding tf-idf weighting, and optimizing current weighting of unigram,bigram,and trigrams. My program is created with an empty stop words array which could be filled with words that we would like to ignore. Even more accuracy improvements could potentially be seen by using a word to vector approach which would allow us to implement different machine learning models altogether such as a Support Vector Machines or Neural Networks.