This Python script will be used to read and analyze a set of PDF files in a given directory as an argument on the command line. The results will be written in the same folder where the script is located. The program will return three types of results. The first one will be keyword clouds created with the abstract of the file, which will be in .png format, located in a directory called /wordClouds and assigned a name that follows this pattern: filename_wordCloud. The second type of result will be a file called results.txt, which will be created in the same folder as the script and will contain the links that each of the analyzed files contains. The last result is a histogram in .png format called histogram_num_figures, which will also be located in the same folder as the script and will show the number of figures each file contains.

It is necessary to have a Grobid server installed correctly and running on your machine, which by default will be running at the address http://localhost:8070. This code has been developed and tested with Python 3.9.12 and pip 21.2.4.

• Build the image

docker build -t <name_dockerFile> .

name_dockerFile the name of the dockerFile

• Run the container

docker run --network=host -v </path/to/my_folder:/path/in/the/container> <name_dockerFile> <path_pdfs>

--network specify the network to be used by the container

-v mount a directory or file from the host machine into the container

name_dockerFile the name of the dockerFile

path_pdfs the path to the directory where the PDFs are located.

Examples:

docker build -t dockerfile
docker  run --network=host -v C:/Users/ramon/IAOS:/app dockerfile /app/pdfs

In case you are not using docker, you can create a conda environment to work with this script, open the terminal and follow the next instructions

• Install Conda: If you don't have Conda installed yet, install it from the official Anaconda website: https://www.anaconda.com/products/distribution
• Create a new environment:

conda create --n <name_enviroment>

• Activate the environment:

conda activate <name_enviroment>

• Install the following dependencies

  conda install grobid_client_python
  

  conda install wordcloud
  

  conda install matplotlib.pyplot
  

  conda install requests
  

  conda install re
  

  conda install os
  

• Execute the program

  python IAOS.py <path_pdfs>
  

When you have finished working in the environment, you can deactivate it with the following command:

conda deactivate 

Distributed under Apache 2.0 license.

Main author and contact: Marta Hidalgo