eth-library-lab/indexda

Natural Language Processing of academic papers for dataset indexing

Welcome 👋

Our mission is to help students, researchers and educators unleash their full potential by boosting ideas that support discovering, accessing, using or sharing scientific information and knowledge. See more on our website.

Projects on Github 💻 📚

We are focussed on bringing projects to a prototype or MVP level at which point they can be shared, utilised and seek support for further development.

Some repos will be ready to reproduce and run end-to-end, others may be in the early experimental development stages. If you are interested in discussing any of our projects please get in touch via our contact form.

Not all of our projects are listed here on Github, so please check out our website for a list of all Library Lab projects

Filsat

repo	eth-library-lab/filsat
about	A transition platform for open source code and online coding tutorials.
tech	Stencil, GraphQL, NodeJS, Django
status	finished 🏁

inDexDa

repo	eth-library-lab/inDexDa
about	Automated identification and indexing of datesets in academic papers using NLP
tech	NLP, webscraping, Tensorflow, BERT,
status	finished 🏁

BioDex

repo	eth-library-lab/biodex--mobile-app
about	Automated species identification of butterflies with a mobile app designed for Natural History Museums
tech	React Native, Django
status	finished 🏁

Image Search

repo	eth-library-lab/open-image-search
about	Image retrieval tool used to search for similar images in an archive and find relevant metadata
tech	Vue, Django, Tensorflow, Docker
status	active 🏃

herbaria--plant-labeling

repo	eth-library-lab/herbaria--plant-labeling
about	Image segmentation model for labelling plant structures in herbarium samples in the family Brassicaceae
tech	Tensorflow, MaskRCNN
status	exploratory 🔬

indexda's Issues

Not able to run InDexDa - "FileNotFoundError"

Hi Parker,

Installation was ok, I was able to follow the documentation and checkout/install the project locally, all good 👍

On the other side, when I try to run the project, with or without tweaking the args.json, I face the following error when I run the command python3 run.py --first_time --scrape --train:

Traceback (most recent call last):
File "run.py", line 48, in
preprocess.processForTrainingBert()
File "/Users/daviddalbusco/projects/lab/datadex/NLP/utils/preprocess.py", line 209, in > processForTrainingBert
self.makeDirs()
File "/Users/daviddalbusco/projects/lab/datadex/NLP/utils/preprocess.py", line 127, in makeDirs
os.mkdir(os.path.join(self.current_dir, '../../data/bert_datatrain/0'))
FileNotFoundError: [Errno 2] No such file or directory: '/Users/daviddalbusco/projects/lab/datadex/NLP/utils/../../data/bert_datatrain/0'

Not rush, definitely something we could try out to solve, if needed, together in January next time we met.

Have a good one
David

doc: missing BERT Network image in README

The BERT Network image referenced in the NLP README is not valid aka no image is displayed