The software has been tested with 9 papers from Google Scholar and works properly. More about that below:
- If one or more PDFs do not have an abstract field the software will fail. Due to this, the software has been updated to check if all the inputs are correct
- The links gotten in the txt file are the ones that are under the tag <p> in the TEI file generated when calling Grobid. There are more which can be found just by looking at the TEI files but no solution has been found yet
As shown here, you can see that a wordcloud has been created for each pdf file given as input
In this image, you can see the bar graphic created by the software. Paper5, which is the nectin_analysis pdf from the image showed above, has only 2 figures that are marked with the <figure> tag in its TEI file
The resulting list of links only shows 1 link from the nectin_analysis pdf