This is the course project repo of Shangyun Lv for W266 2019 Fall This is a one-person group project all the way through, all the code and write-up is done by Shangyun himself.
The project uses deep learning methods to see whether people with different profiles demonstrate different writing styles; if they do, how they are different and what kind of pattern we could find.
The project uses the data provided in The Blog Authorship Corpus for the project, which incorporates a total of 681,288 blog posts (over 140 million words) of 19,320 bloggers gathered from blogger.com in August 2004 - or approximately 35 posts and 7250 words per person. The features of the model will be generated through the blog post raw text in .xml format. The four labels provided as file name in the corpus includes gender, age, industry, and astrological sign of the writer.
The notebook is ordered based on the process of the text pre-processing and model training.
The link to course report google doc: https://docs.google.com/document/d/1eFeUiXQ9xCOtoopr5rKnIOET7HeTk84jCu2PqHC4sAE/edit?usp=sharing