Vis-Seive

This project aims to enable visualization service providers/facilitators to get a sense of the kinds of visualizations that are being produced by their institution as a whole. This provides a range of benefits

• Informed Facility Development
• Trend tracking
• Technique discovery/development
• Service catalog entry inspiration

The overall project covers

• the gathering of publication information
• the retrieval of open access pdfs
• the scraping of figures from pdfs
• the aggregation of figures into a largescale mosaic grouped by field and style

Members

• Carolina Roe-Raymond
• Devin Bayly
• Ben Kruse

Label studio

Setting up

Install it with pip

pip install label-studio

For locally hosted files, set the following environment variables. Note: this assumes you are in the vis-siev directory, otherwise set the LOCAL_FILES_DOCUMENT_ROOT to whatevers one above your content folder.

export LOCAL_FILES_SERVING_ENABLED=true
export LABEL_STUDIO_LOCAL_FILES_DOCUMENT_ROOT=$(pwd)/database

Then run the server

label-studio

Creating a new project

Go the local host address label studio gives you after the last step.

Create a new project, adding the template code. Don't import the data here, well do that later.

Importing data

Go to settings->cloud storage->Add Source Storage then select Local Files as the storage type. For absolute local path, put the absolute path to the database folder (for me: /home/benk/Documents/datavis/vis-sieve/database/content). Set the regex filter to .*jpg and enable Treat every bucket object as a source file. Then click Add Storage (or check the connection first).

This sets up the source, to actually bring in the data, click sync storage in the cloud storage tab (this may take a second).