A tool for the manual annotating linguistic data

with the Best Worst Scaling method

Create new projects

Choose a phenomenon to analyze,

load your instance file

and you are ready to start!

Share your project

Authorizes other users to

annotate your project

by indicating their email

Get a gold standard

Generate a gold dataset

that represent the truth against

wich comparing futures predictions

Best–Worst Scaling

Best–Worst Scaling (BWS) is a less-known, and more recently introduced, variant of comparative annotation. Annotators are presented with n items at a time (an n-tuple, where n > 1, and typically n = 4). They are asked which item is the best (highest in terms of the property of interest) and which is the worst (lowest in terms of the property of interest). How tuples are created is better explained here. The scores can be used to rank items by the property of interest.

My name is Cagnazzo Christian and I am a third year student of the Triennial of computer science at the University of Turin. This project was born thanks to professor Valerio Basile as a thesis proposal, of which he is the speaker. I thank the professor for providing me with his initial library that implemented an interactive interface for Best-worst Scaling annotation for a single user, which I then expanded, made multiuser with a database and used to create a Restful API for this tool. In case of a bug write to: christiancagnazzo@hotmail.com