Shan/Final Project
From LING073
Code is here: https://github.com/kanyarinb/ling073-shn
Initial Evaluation
coverage: 1296546 / 1943840 (~0.66700242818338957939)
remaining unknown forms: 647294
ambiguity: 2519646 / 1943840 (~1.29622088237715038275)
ambiguity with disambiguator: 2420618 / 1943840 (~1.24527636019425467117)
Precision: 68.87417%
Recall: 61.90476%
Evaluation after Changes
coverage: 1704213 / 2246313 (~0.75867120922151098266)
remaining unknown forms: 542100
ambiguity: 2952314 / 2246313 (~1.31429324408486261710)
ambiguity with disambiguator: 2695226 / 2246313 (~1.19984436719192739391)
Precision: 71.94719%
Recall: 64.30678%