lpuettmann / patent-automat

Search through corpus of patent texts to measure automation :computer: over time and its effect.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Analyze Patent Grant Texts

Katja Mann, Lukas Püttmann

Overview:

This repository holds the codes for our paper. Please let us know if you find any bugs by filing an issue or sending us an email. You are free to use them (see license). If you do, please cite us.

Get data:

Use these codes to obtain the patent grant texts.

Before running:

In directory /specs in in set_data_path.m specify the absolute paths to the patent text data.

Run set_up.m to add the current directory and all underlying directories to Matlab's search path.

Only then run testPatentAutomat.m to use Matlab's xUnit testing framework which checks if the code functions properly. There should be no warnings or error messages here.

To run:

Run run_patentsearch.m. You can find all the figures and tables that go into the paper in the directory /output.

Outputs:

The program creates among other things sic_automix_allyears.csv. This industry-year dataset is described in more details here.

Necessary software:

We wrote and tested this in Matlab R2013b on Windows, OS X and Linux machines. No toolboxes are used. The tests rely on Matlab's xUnit Framework which are included in Matlab from version R2013a (March 2013) onwards or else you downloaded them here.

About

Search through corpus of patent texts to measure automation :computer: over time and its effect.

License:MIT License


Languages

Language:MATLAB 92.7%Language:R 6.7%Language:Stata 0.5%Language:Objective-C 0.1%