This repository contains the modifications of code from Chuan Wang's CAMR parser to incorporate 3 types of world knowledge sources.
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
Aligner.py
README
amr_parsing.py
class_amr_analysis.py
constants.py
depparser.py
gen_crf_files.py
gen_crf_for_all.py
gen_crf_for_all_NE.py
gen_crf_with_name_children_files.py
gen_train_ner_data.py
get_entity_links.py
get_general_AMR_features.py
get_has_name.py
get_ner_amr_features.py
get_roget_classes.py
get_wordnet_features.py
get_wordnet_size.py
graphstate.py
model.py
newstate.py
oracle.py
parser.py
perceptron.py
preprocessing.py
profiling.py
roget_class.py
span.py
test_vload.py
wordnet_features.py

README

This is where I will put the code for LREC paper:

Welch, Charlie, Kummerfeld, Jonathan K, Feng, Song, & Mihalcea, Rada (2018). World Knowledge for Semantic Parsing with Abstract Meaning Representation. In 11th Language Resources and Evaluation Conference (LREC). 


Commands for running CAMR

To retrain the model:

python amr_parsing.py -m train --amrfmt amr --verblist --smatcheval --model amr_model0 --feat ./feature/basic_abt_brown_feats.templates train.txt -d dev.txt > ./logs/train.log 2>&1 &

To parse the test set:

python amr_parsing.py -m parse --model amr_model0.m test-sentences.txt 2>logs/error.log

To run basic SMATCH evaluation:

python smatch_2.0.2/smatch.py -f test-sentences.txt.all.amr_model0.parsed gold_LDC2014T12 --pr

To run detailed SMATCH:

cd amr-evaluation
./evaluation.sh ../test-sentences.txt.all.amr_model0.parsed ../gold_LDC2014T12