ProFluency-Network (Processing Fluency Neural Network)

Below I give a short summary of the contents and the most important files

📁 Codes

📁 controls

The variables for the panel regressions are stored here.
These include the control variables from Compustat.

📁 data

Raw data that are either directly from the source or used to give extra insights.

data_nyse_screener.xlsx contains the (cleaned) company names from the NASDAQ Screener.
data_lemma_internet.txt is the frequency list, note that you to multiply the 'frequency' with 181,376 to get the real total amount of occurences.
data_lexicon.xlsx is the lexicon, the list that contains many English words.

📁 neural network

This file contains the architecture of the neural network.

📁 output neural network

z_predictions_100x.xlsx is the file that contains the fluency predictions for each company name.

📁 panel regressions

Contains the 3 panel regressions in Eviews.

📁 python

Contains Jupyter Notebook files written in python.

LaTeX

📁 figures

Contains the figures that are used in this project.

🗒️ library.bib

Bibtex style references to other studies.

🗒️ main.tex

This file contains the thesis written in LaTeX.

Prefixes

If you were curious why... 💭

• data_                  Raw data obtained from scources

• data_filtered_   Processed raw data

• data_reg_           Regression data for testing the effect of fluency on the characteristics

• z_                       Input and output data that are from the deep learning model

• py_                     Jupyter Notebook files with python code