def get_data(self):
process_wmt = PrepareWmt()
data_structure1 = process_wmt.get_data_structure(self.config)
data_structure2 = process_wmt.get_data_structure2(self.config)
process_wmt.print_data_set(self.config, data_structure1)
if 'Parse' in loads(self.config.get("Resources", "processors")):
process_wmt_parse = PrepareWmt(data_type='parse')
data_structure_parse = process_wmt_parse.get_data_structure(self.config)
process_wmt_parse.print_data_set(self.config, data_structure_parse)
f_judgements = self.config.get('WMT', 'human_ranking')
maximum_comparisons = int(self.config.get('WMT', 'maximum_comparisons'))
human_rankings = HumanRanking()
human_rankings.add_human_data(f_judgements, self.config, max_comparisons=maximum_comparisons)
process = Process(self.config)
sents_tgt, sents_ref = process.run_processors()
extractor = FeatureExtractor(self.config)
features_to_extract = FeatureExtractor.read_feature_names(self.config)
extractor.extract_features(features_to_extract, sents_tgt, sents_ref)
return data_structure2, human_rankings, extractor.vals
def feature_extraction(config_features_path):
config = ConfigParser()
config.readfp(open(config_features_path))
wd = config.get('WMT', 'working_directory')
if not os.path.exists(wd):
os.mkdir(wd)
data = RankingData(config)
data.read_dataset()
process = Process(config)
sentences_tgt, sentences_ref = process.run_processors()
feature_names = FeatureExtractor.read_feature_names(config)
feature_values = FeatureExtractor.extract_features_static(feature_names, sentences_tgt, sentences_ref)
write_feature_file(wd + '/' + 'x' + '_' + data.datasets[0].name + '.tsv', feature_values)
my_dataset = data.plain[0].dataset
my_lp = data.plain[0].lp
f_path = wd + '/' + 'x' + '_' + my_dataset + '_' + my_lp + '.tsv'
f_file = open(f_path, 'w')
for i, instance in enumerate(data.plain):
if instance.dataset == my_dataset and instance.lp == my_lp:
f_file.write('\t'.join([str(x) for x in feature_values[i]]) + "\n")
else:
f_file.close()
my_dataset = instance.dataset
my_lp = instance.lp
f_path = wd + '/' + 'x' + '_' + my_dataset + '_' + my_lp + '.tsv'
f_file = open(f_path, 'w')
f_judgements = config.get('WMT', 'human_ranking')
human_rankings = HumanRanking()
human_rankings.add_human_data(f_judgements, config)
human_rankings.get_sentence_ids(data)
learn_to_rank(feature_values, human_rankings, wd + '/' + 'x_learn_to_rank.tsv', wd + '/' + 'y_learn_to_rank.tsv')
def clean_dataset(config_learning, human_comparisons):
feature_values = read_features_file(config_learning.get('x_train'), '\t')
labels = read_reference_file(config_learning.get('y_train'), '\t')
new_feature_values = []
new_labels = []
human_comparisons = RankingTask.eliminate_ties(human_comparisons)
comparisons_untied_phrases = defaultdict(list)
comparisons_untied_signs = defaultdict(list)
deduplicated_phrases, deduplicated_signs = HumanRanking.deduplicate(human_comparisons)
for dataset, lang_pair in sorted(human_comparisons.keys()):
for comparison in human_comparisons[dataset, lang_pair]:
if comparison.sign == "=":
continue
else:
comparisons_untied_phrases[dataset, lang_pair].append([comparison.phrase, comparison.sys1, comparison.sys2])
comparisons_untied_signs[dataset, lang_pair].append(comparison.sign)
for dataset, lang_pair in sorted(human_comparisons.keys()):
for i, comparison in enumerate(comparisons_untied_phrases[dataset, lang_pair]):
features = feature_values[i]
label = labels[i]
if comparison in deduplicated_phrases[dataset, lang_pair]:
if deduplicated_signs[dataset, lang_pair][deduplicated_phrases[dataset, lang_pair].index(comparison)] is None:
continue
label = RankingTask.signs_to_labels(deduplicated_signs[dataset, lang_pair][deduplicated_phrases[dataset, lang_pair].index(comparison)])
new_feature_values.append(features)
new_labels.append(label)
write_feature_file(config_learning.get('x_train') + "." + "clean", new_feature_values)
write_reference_file(config_learning.get('y_train') + "." + "clean", new_labels)
config_learning = yaml.load(cfg_file.read())
# Prepare feature files
# This needs to be done for both training and testing data, changing the names of the datasets in the configuratio file
prepare_wmt = PrepareWmt()
ranking_task = RankingTask(config_path)
ranking_task.prepare_feature_files()
# Create training set for learn to rank
# Comment the above prepare feature files method
dataset_for_all = config.get('WMT', 'dataset')
feature_set_name = os.path.basename(config.get('Features', 'feature_set')).replace(".txt", "")
data_structure2 = prepare_wmt.get_data_structure2(config)
f_judgements = config.get('WMT', 'human_ranking')
human_rankings = HumanRanking()
human_rankings.add_human_data(f_judgements, config)
feature_values = read_features_file(os.path.expanduser(config.get('WMT', 'output_dir')) + '/' + 'x_' + dataset_for_all + '.' + feature_set_name + '.' + 'all' + '.tsv', "\t")
ranking_task.training_set_for_learn_to_rank(data_structure2, human_rankings, feature_values)
ranking_task.train_save(config_learning, config)
# Run the trained model on a the test feature file and produce the output in WMT format
predictions = ranking_task.test_learn_to_rank_coefficients(config_learning, config)
data_structure = prepare_wmt.get_data_structure(config)
prepare_wmt.wmt_format(config, feature_set_name, dataset_for_all, predictions, data_structure)
