Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/vtlim/patfam

Web app to determine whether patent applications from different jurisdictions (USPTO, EPO, WIPO, etc.) are of the same family.
https://github.com/vtlim/patfam

css d3js epo html javascript materializecss patent-applications patent-offices python selenium uspto web-scraping wipo

Last synced: 17 days ago
JSON representation

Web app to determine whether patent applications from different jurisdictions (USPTO, EPO, WIPO, etc.) are of the same family.

Awesome Lists containing this project

README 

        
# PatFam: Patent Families



_README last edited_: 16 Dec 2020



The purpose of this tool is to allow users to easily determine whether patent applications are in the same family. Instead of having to check each individual patent office's website, users can enter patent, application, or publication numbers for various patent offices and see the relationship between the documents. 



Currently supports: TBD



![screenshot](https://github.com/vtlim/patfam/blob/foundations/archive/screenshot.png)



## Setup



### Summary of conda installs

See below for important notes including system dependencies, version compatibility, access keys, etc.

```

conda create -n patfam python=3.6

conda activate patfam



conda install pip

pip install uspto-opendata-python



conda install -c anaconda lxml networkx pygraphviz graphviz flask gunicorn

conda install -c conda-forge selenium matplotlib

```



### [USPTO] [USPTO Open Data API Client](https://docs.ip-tools.org/uspto-opendata-python/index.html)

For data from the United States Patent and Trademark Office (USPTO), we'll use the API client for the USPTO Patent Examination Data System (PEDS). Do not mix up PEDS with PBD, which is the USPTO PAIR Bulk Data (PBD) system. The PBD has been decommissioned.



```

# install libxml2 dependencies (make sure you have -dev versions)

sudo apt-get update

sudo apt-get install libxml2-dev libxslt1-dev python-lxml



# create conda environment and install some packages (see note below)

conda create -n patfam python=3.6

conda activate patfam

conda install -c anaconda lxml



# install pip (if you don't already have it) and the api client

conda install pip

pip install uspto-opendata-python



# run a test query

uspto-peds get "15431686" --type=application --format=xml

```

NOTE: As of Dec. 2020, `uspto-opendata-python` will not install if using Python 3.9.0 (returning gcc compilation error of [this type](https://github.com/pandas-dev/pandas/issues/32114)).



#### See my tutorial walkthrough of the USPTO Open Data API Client [here](uspto/explore_uspto_data.ipynb).



### [EPO] [Python EPO OPS Client](https://github.com/gsong/python-epo-ops-client)

For data from the European Patent Office (EPO), we'll use the Open Patent Services (OPS) client developed by George Song et al. In order to get API access, you will need to [request access credentials](https://developers.epo.org/) from OPS. After I submitted my request, I was granted access from EPO the next day.

```

conda activate patfam

pip install python-epo-ops-client

```

#### See my tutorial walkthrough of the Python EPO OPS Client [here](epo/explore_epo_data.ipynb).



### Web Scraping



As far as I can tell, there is no freely available API access to the World Intellectual Property Organization (WIPO; for PCT applications) or to the Japan Patent Office (JPO). For those applications, we'll request data from the web directly.



Before this I tried the Python `requests` package to obtain site data. I made a GET search query to WIPO PatentScope for `docId=WO2001029057`. However, no patent-related data could be found in the HTML content. It looks like the site is rendered dynamically using JavaScript so we'll use Selenium instead:



```

conda activate patfam

conda install -c conda-forge selenium

```



Since I'm working in the Windows Subsystem for Linux (WSL2), I needed to download an Internet browser and related driver (following the [tutorial](https://www.gregbrisebois.com/posts/chromedriver-in-wsl2/) from Greg Brisebois). If you already have an Internet browser in the same OS that you're working in, you would just get to obtain the relevant driver for your browser, browser version, and OS.



```

# install dependencies

sudo apt-get update

sudo apt-get install -y curl unzip xvfb libxi6 libgconf-2-4



# get chrome browser

wget https://dl.google.com/linux/direct/google-chrome-stable_current_amd64.deb

sudo apt install ./google-chrome-stable_current_amd64.deb



# optional - prevent updates of google chrome (if it updates, you also have to update the chromedriver)

sudo apt-mark hold google-chrome-stable



# check browser install

google-chrome --version



# get chrome driver; USE THE DRIVER VERSION RELEVANT TO YOUR SETUP

wget https://chromedriver.storage.googleapis.com/87.0.4280.88/chromedriver_linux64.zip

unzip chromedriver_linux64.zip

sudo mv chromedriver /usr/bin/chromedriver

sudo chown root:root /usr/bin/chromedriver

sudo chmod +x /usr/bin/chromedriver



# check driver install

chromedriver --version

```



Optionally, if you want to work with the Chrome window (as opposed to using it in "headless form"), try the command `google-chrome` to check that the window will come up. I had to resolve a few issues with WSL2 on my end before it worked. More details [here](https://github.com/vtlim/patfam/blob/main/wsl2_xserver.md).



#### See an example of using Selenium to obtain WIPO PatentScope data [here](wipo/explore_wipo_data.ipynb).

#### See an example of using Selenium to obtain JPO patent data [here](jpo/explore_jpo_data.ipynb).