Skip to content

The code implements the BLLIP metric which is an automatic machine translation evaluation metric that uses dependency sets for comparing reference and candidate sentences.

Notifications You must be signed in to change notification settings

MrudulaSatya/BLLIP

Repository files navigation

BLLIP

The code implements the BLLIP metric which is an automatic machine translation evaluation metric that uses dependency sets for comparing reference and candidate sentences.

This code uses the Hindi Dependency Parser designed by Siva Reddy to parse the candidate and reference sentences and get the dependency tags for them in two different output files (one for candidate and anither for reference sentences). Source: http://sivareddy.in/downloads

The reference paper used is: Pozar, Michael & Charniak, Eugene. (2018). Bllip: An Improved Evaluation Metric for Machine Translation.

About

The code implements the BLLIP metric which is an automatic machine translation evaluation metric that uses dependency sets for comparing reference and candidate sentences.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages