def __init__(self, labels_scaler, stemmer, frames_version, model_io,
terms_per_context, dist_in_terms_between_attitude_ends):
assert (isinstance(frames_version, RuSentiFramesVersions)
or frames_version is None)
assert (isinstance(model_io, BertModelIO))
assert (isinstance(terms_per_context, int))
self.__dist_in_terms_between_attitude_ends = dist_in_terms_between_attitude_ends
super(CustomSerializationData, self).__init__(labels_scaler, stemmer)
self.__model_io = model_io
self.__terms_per_context = terms_per_context
self.__opinion_formatter = RuSentRelOpinionCollectionFormatter()
self.__frames_collection = None
self.__unique_frame_variants = None
if frames_version is not None:
self.__frames_collection = RuSentiFramesCollection.read_collection(
version=frames_version)
self.__unique_frame_variants = FrameVariantsCollection.create_unique_variants_from_iterable(
variants_with_id=self.__frames_collection.
iter_frame_id_and_variants(),
stemmer=stemmer)
def about_version(version=RuSentiFramesVersions.V20):
stemmer = MystemWrapper()
pos_tagger = POSMystemWrapper(stemmer.MystemInstance)
frames_collection = RuSentiFramesCollection.read_collection(
version=version)
print("Lexicon version:", version)
return __about(frames_collection=frames_collection, pos_tagger=pos_tagger)
def __init__(self, labels_scaler, stemmer, pos_tagger, embedding,
dist_in_terms_between_att_ends, terms_per_context,
frames_version, str_entity_formatter, opinion_formatter,
rusentrel_version):
assert (isinstance(embedding, Embedding))
assert (isinstance(stemmer, Stemmer))
assert (isinstance(pos_tagger, POSTagger))
assert (isinstance(rusentrel_version, RuSentRelVersions))
assert (isinstance(frames_version, RuSentiFramesVersions))
assert (isinstance(str_entity_formatter, StringEntitiesFormatter))
assert (isinstance(opinion_formatter, OpinionCollectionsFormatter))
assert (isinstance(dist_in_terms_between_att_ends, int)
or dist_in_terms_between_att_ends is None)
assert (isinstance(terms_per_context, int))
self.__dist_in_terms_between_att_ends = dist_in_terms_between_att_ends
super(RuSentRelExperimentSerializationData,
self).__init__(labels_scaler=labels_scaler, stemmer=stemmer)
self.__pos_tagger = pos_tagger
self.__terms_per_context = terms_per_context
self.__rusentrel_version = rusentrel_version
self.__str_entity_formatter = str_entity_formatter
self.__word_embedding = embedding
self.__opinion_formatter = opinion_formatter
self.__frames_collection = RuSentiFramesCollection.read_collection(
version=frames_version)
self.__unique_frame_variants = FrameVariantsCollection.create_unique_variants_from_iterable(
variants_with_id=self.__frames_collection.
iter_frame_id_and_variants(),
stemmer=self.Stemmer)
def __iter_frame_variants():
frames_collection = RuSentiFramesCollection.read_collection(RuSentiFramesVersions.V20)
frame_variants = FrameVariantsCollection()
frame_variants.fill_from_iterable(variants_with_id=frames_collection.iter_frame_id_and_variants(),
overwrite_existed_variant=True,
raise_error_on_existed_variant=False)
for v, _ in frame_variants.iter_variants():
yield v
def __create_frames_variants_collection(self):
frames = RuSentiFramesCollection.read_collection(
RuSentiFramesVersions.V20)
frame_variant_collection = FrameVariantsCollection()
frame_variant_collection.fill_from_iterable(
variants_with_id=frames.iter_frame_id_and_variants(),
overwrite_existed_variant=True,
raise_error_on_existed_variant=False)
return frame_variant_collection
def test_reading(self):
# Initializing logger
logger = logging.getLogger(__name__)
logger.setLevel(logging.INFO)
logging.basicConfig(level=logging.DEBUG)
frames = RuSentiFramesCollection.read_collection(
RuSentiFramesVersions.V20)
for frame_id in frames.iter_frames_ids():
# id
logger.info("Frame: {}".format(frame_id))
# titles
logger.info("Titles: {}".format(",".join(
frames.get_frame_titles(frame_id))))
# variants
logger.info("Variants: {}".format(",".join(
frames.get_frame_variants(frame_id))))
# roles
for role in frames.get_frame_roles(frame_id):
logger.info("Role: {}".format(" -- ".join(
[role.Source, role.Description])))
# states
for state in frames.get_frame_states(frame_id):
logger.info("State: {}".format(",".join(
[state.Role,
state.Label.to_class_str(),
str(state.Prob)])))
# polarity
for polarity in frames.get_frame_polarities(frame_id):
logger.info("Polarity: {}".format(",".join([
polarity.Source, polarity.Destination,
polarity.Label.to_class_str()
])))
has_a0_a1_pol = frames.try_get_frame_polarity(frame_id,
role_src="a0",
role_dest="a1")
logger.info(
"Has a0->a1 polarity: {}".format(has_a0_a1_pol is not None))
# frame variants.
frame_variants = FrameVariantsCollection()
frame_variants.fill_from_iterable(
variants_with_id=frames.iter_frame_id_and_variants(),
overwrite_existed_variant=True,
raise_error_on_existed_variant=False)
frame_variant = frame_variants.get_variant_by_value("хвалить")
assert (isinstance(frame_variant, FrameVariant))
logger.info("FrameVariantValue: {}".format(frame_variant.get_value()))
logger.info("FrameID: {}".format(frame_variant.FrameID))
def setUpClass(cls):
cls.stemmer = MystemWrapper()
cls.entities_formatter = RussianEntitiesCasedFormatter(
pos_tagger=POSMystemWrapper(Mystem(entire_input=False)))
cls.synonyms = RuSentRelSynonymsCollectionProvider.load_collection(
stemmer=cls.stemmer)
cls.frames_collection = RuSentiFramesCollection.read_collection(
version=RuSentiFramesVersions.V10)
cls.unique_frame_variants = FrameVariantsCollection()
cls.unique_frame_variants.fill_from_iterable(
variants_with_id=cls.frames_collection.iter_frame_id_and_variants(
),
overwrite_existed_variant=True,
raise_error_on_existed_variant=False)
def test_parsing(self):
# Initializing logger.
logger = logging.getLogger(__name__)
logger.setLevel(logging.DEBUG)
logging.basicConfig(level=logging.DEBUG)
# Initializing stemmer.
stemmer = MystemWrapper()
# frame and variants.
frames = RuSentiFramesCollection.read_collection(version=RuSentiFramesVersions.V20)
frame_variants = FrameVariantsCollection()
frame_variants.fill_from_iterable(variants_with_id=frames.iter_frame_id_and_variants(),
overwrite_existed_variant=True,
raise_error_on_existed_variant=False)
text_parser = BaseTextParser(pipeline=[RuSentRelTextEntitiesParser(),
DefaultTextTokenizer(keep_tokens=True),
LemmasBasedFrameVariantsParser(frame_variants=frame_variants,
stemmer=stemmer,
save_lemmas=False),
FrameVariantsSentimentNegation()])
# Reading synonyms collection.
synonyms = RuSentRelSynonymsCollectionProvider.load_collection(stemmer=stemmer)
version = RuSentRelVersions.V11
for doc_id in RuSentRelIOUtils.iter_collection_indices(version):
# Parsing
news = RuSentRelNews.read_document(doc_id=doc_id,
synonyms=synonyms,
version=version)
# Perform text parsing.
parsed_news = NewsParser.parse(news=news, text_parser=text_parser)
debug_show_news_terms(parsed_news=parsed_news)
def create_frames_collection():
return RuSentiFramesCollection.read_collection(
version=RuSentiFramesVersions.V20,
labels_fmt=ExperimentRuSentiFramesLabelsFormatter())