def get_final_semeval_data(classes, train_loc, dev_loc, test_loc):
"""
Loads the final version of the semeval data that was merged by the following criteria:
train: train13, dev13, train14, train15
dev: test13, dev14, dev15
test: test14, test15
It should be noted that some redundancy exists between each year's set
:param classes: bitflags following the definitions in preprocessing
:param train_loc: location of training data
:param dev_loc: location of development data
:param test_loc: location of test data
:return: tuple of lists, following the (X, y) scheme
"""
train_labels, train_tweets, train_pos = parse(
train_loc, classes
)
dev_labels, dev_tweets, dev_pos = parse(
dev_loc, classes
)
test_labels, test_tweets, test_pos = parse(
test_loc, classes
)
train = [e[0]+'\t'+e[1] for e in zip(train_tweets, train_pos)], train_labels
dev = [e[0]+'\t'+e[1] for e in zip(dev_tweets, dev_pos)], dev_labels
test = [e[0]+'\t'+e[1] for e in zip(test_tweets, test_pos)], test_labels
return train, dev, test
def re_train(model, model_loc, train, dev, corpus_loc, num, num_runs, classes):
"""
runs the back feed mechanism.
:param model: classificator
:param model_loc: If the model was not loaded, a raw one is trained and
saved to this location
:param train: File with annotated training tweets
:param dev: File with annotated development tweets
:param corpus_loc: Directory of the Sentiment140 corpus
:param num: Amount of added corpus data per iteration
:param num_runs: Number of iterations
:param classes: Variable holding bit flags corresponding to class labels --
definition is in preprocessing variable 'ltd'
"""
# load training data, is needed either way
train_labels, train_tweets, train_pos = preprocessing.parse(train, classes)
x_train = [e[0] + "\t" + e[1] for e in zip(train_tweets, train_pos)]
y_train = train_labels
# x_train, y_train = uniformify(x_train, y_train)
# if model is raw, train it
if not hasattr(model, "grid_scores_"):
logging.info("No trained model given, building training features for " "binary class codes: %s" % bin(classes))
model.fit(x_train, y_train)
logging.info("writing new model to disk at %s.." % model_loc)
with open(model_loc, "w") as sink:
cPickle.dump(model, sink)
logging.info("done.")
# get test data
dev_labels, dev_tweets, dev_pos = preprocessing.parse(dev, classes)
x_dev = [e[0] + "\t" + e[1] for e in zip(dev_tweets, dev_pos)]
y_dev = dev_labels
# x_dev, y_dev = uniformify(x_dev, y_dev)
# initial eval
logging.info("Initial evaluation..")
print_scores(model, x_dev, y_dev, classes)
# print label distribution, in order to check on how the ratio of pos to neg
# influences the scoring. Seems pretty balanced now that len(t_gold_1) is
# equal to len(t_gold_2)
# print('num t_gold_1 '+str(len([l for l in y_train if l == 1])))
# print('num t_gold_2 '+str(len([l for l in y_train if l == 2])))
# print('t_1 '+str(len([l for l in model.predict(x_train) if l == 1])))
# print('t_2 '+str(len([l for l in model.predict(x_train) if l == 2])))
# print('num d_gold_1 '+str(len([l for l in y_dev if l == 1])))
# print('num d_gold_2 '+str(len([l for l in y_dev if l == 2])))
# print('d_1 '+str(len([l for l in model.predict(x_dev) if l == 1])))
# print('d_2 '+str(len([l for l in model.predict(x_dev) if l == 2])))
# feedback loop
logging.info("Initializing backfeed instance..")
feed = Feeder(corpus_loc)
logging.info("done. Now starting backfeed loop")
for count in range(1, num_runs + 1):
feed.add_best_n(model, num, x_train, y_train)
logging.info("Retrain run %i" % count)
print_scores(model, x_dev, y_dev, classes)
def test(sc, file_positive, files_negative, file_model):
"""
Tests a classification model using positive samples in file_positive and
negative samples in file_negative. It prints the results to standard output
:param sc: The spark context
:type sc: SparkContext
:param file_positive: The file with tweets to predict
:type file_positive: str
:param files_negative: The files with tweets to reject
:type files_negative: list[str]
:param file_model: The file where the model is located
:type file_model: str
"""
tweets_positive = sc.textFile(file_positive).map(parse_json).filter(lambda x: is_valid(x) and is_english(x)).cache()
list_negatives = [sc.textFile(file_negative).map(parse_json).filter(lambda x: is_valid(x) and is_english(x)) for file_negative in files_negative]
tweets_negative = list_negatives[0]
for ln in list_negatives[1:]:
tweets_negative = tweets_negative.union(ln)
try:
print("Reading stored classification model")
model = pickle.load(open(file_model, 'rb'))
print("Computing predictions")
threshold = 0.0
total_positive = tweets_positive.count()
total_negative = tweets_negative.count()
true_positives = tweets_positive.filter(lambda x: model.predict(parse(x)) > threshold).count()
true_negatives = tweets_negative.filter(lambda x: model.predict(parse(x)) <= threshold).count()
false_negatives = total_positive - true_positives
false_positives = total_negative - true_negatives
print("Results for %s:" % file_model)
print(" Total positives: %d" % total_positive)
print(" Total negatives: %d" % total_negative)
print(" False positives: %d" % false_positives)
print(" False negatives: %d" % false_negatives)
precision = 0.0
recall = 0.0
try:
precision = float(true_positives) / float(true_positives + false_positives)
recall = float(true_positives) / float(true_positives + false_negatives)
except:
pass
print(" Precision: %f" % precision)
print(" Recall: %f" % recall)
print("Done!")
except Exception as e:
print("Error:")
print(e)
def get_train_data():
classes = POS | NEU | NEG
train_loc = root+'Data/twitterData/train_alternative.tsv'
dev_loc = root+'Data/twitterData/dev_alternative.tsv'
test_loc = root+'Data/twitterData/test_alternative.tsv'
train_labels, train_tweets, train_pos = parse(
train_loc, classes
)
dev_labels, dev_tweets, dev_pos = parse(
dev_loc, classes
)
test_labels, test_tweets, test_pos = parse(
test_loc, classes
)
semeval_train = [e[0]+'\t'+e[1] for e in zip(train_tweets, train_pos)], train_labels
semeval_dev = [e[0]+'\t'+e[1] for e in zip(dev_tweets, dev_pos)], dev_labels
semeval_test = [e[0]+'\t'+e[1] for e in zip(test_tweets, test_pos)], test_labels
return semeval_train, semeval_dev, semeval_test
def predict_labels(tweet):
x = parse(tweet)
labels = []
if sports.predict(x) > 0.0:
labels.append("sports")
if politics.predict(x) > 0.0:
labels.append("politics")
if technology.predict(x):
labels.append("technology")
return labels
def predict_labels(tweet):
x = parse(tweet)
labels = []
if sports.predict(x) > 0.0:
labels.append("sports")
if politics.predict(x) > 0.0:
labels.append("politics")
if technology.predict(x):
labels.append("technology")
return labels
def score(model, model_loc, train, dev, classes):
"""
Trains and tests the model with the current feature extractors and
parameters.
:param model: Model to be tested, either instance of gird_search, or a path
to a previously saved model
:param model_loc: If the model was not loaded, a raw one is trained and
saved to this location
:param train: File with annotated training tweets
:param dev: File with annotated development tweets
:param classes: Amount of classes which get drawn from train and test sets
:return: best parameters found by the current grid search
"""
# if model is raw, train it
if not hasattr(model, "grid_scores_"):
logging.info("No trained model given, building training features for " "binary class codes: %s" % bin(classes))
train_labels, train_tweets, train_pos = preprocessing.parse(train, classes)
x_train = [e[0] + "\t" + e[1] for e in zip(train_tweets, train_pos)]
y_train = train_labels
# x_train, y_train = uniformify(x_train, y_train)
# [u'#RonPaul campaign manager Doug Wead : Contrary to media lies , doing JUST GREAT with delegates even after Super Tuesday http://url\t# N N ^ ^ , A P N N , V R A P N R P ^ ^ U']
model.fit(x_train, y_train)
logging.info("writing new model to disk at %s.." % model_loc)
with open(model_loc, "w") as sink:
cPickle.dump(model, sink)
logging.info("done.")
# get test data
dev_labels, dev_tweets, dev_pos = preprocessing.parse(dev, classes)
x_dev = [e[0] + "\t" + e[1] for e in zip(dev_tweets, dev_pos)]
y_dev = dev_labels
# x_dev, y_dev = uniformify(x_dev, y_dev)
# test and evaluate
print_scores(model, x_dev, y_dev, classes)
return model.best_params_
"""This module produces SemEval Task 3, Subtask B datasets in JSON."""
from itertools import chain
from json import dump
from preprocessing import parse
from xmlfiles import XMLFiles
if __name__ == "__main__":
result = {}
for year, dataset in [(2016, "train"), (2016, "dev"), (2016, "test"),
(2017, "test")]:
with XMLFiles(year, dataset) as xmlfiles:
result["%d-%s" % (year, dataset)] = list(
chain(*[parse(xmlfile) for xmlfile in xmlfiles]))
with open("result.json", "wb") as jsonfile:
dump(result, jsonfile, sort_keys=True, indent=4)
"""This module produces unannotated SemEval Task 3, Subtask A datasets in JSON."""
from itertools import chain
from json import dump
from preprocessing import parse
from xmlfiles import XMLFiles
if __name__ == "__main__":
result = []
with XMLFiles(2016, "unannotated") as xmlfiles:
assert len(xmlfiles) == 1
corpus = parse(xmlfiles[0])
with open("result.json", "w") as jsonfile:
dump(corpus, jsonfile, sort_keys=True, indent=4)
import time
import pandas as pd
from preprocessing import parse
from app import appriori
from helper import printResult
from fp import FPGrowth
from ImprovedApp import PartitionedApp
start = time.time()
df = pd.read_csv('adult.data.csv', names=[0,1,2,3,4,5,6,7,8,9,10,11,12,13,14])
dataSet= parse(df, df.shape[0])
sec = time.time() - start
print("\nData Preprocessing time: " + str(sec)+" seconds")
minSup = int(input("Enter minimun relative support as a percentage (0 - 100) : "))
print('\n--------------------------------- APRIORI --------------------------------------')
#
#start = time.time()
#feqApp = appriori(dataSet,minSup)
#printResult(feqApp)
#sec = time.time() - start
#print("\nApriori Execution time: " + str(sec)+" seconds")
#print('\n--------------------------------- FP GROWTH --------------------------------------')
start = time.time()
feqFP = FPGrowth(dataSet,minSup)
printResult(feqFP)
sec = time.time() - start
print("\nFP Growth Execution time: " + str(sec)+" seconds")
print('\n--------------------------------- PARTITIONED APRIORI --------------------------------')
#start = time.time()
"/media/gaurish/Angela's Files/projects/semeval-2016_2017-task3-subtaskA-unannotated-english/v3.2/train/SemEval2016-Task3-CQA-QL-train-part2-subtaskA.xml",
"/media/gaurish/Angela's Files/projects/semeval-2016_2017-task3-subtaskA-unannotated-english/v3.2/dev/SemEval2016-Task3-CQA-QL-dev-subtaskA.xml"
]
counter =[]
def write_to_file(data,output_path,mode ="w"):
with open(output_path, mode, encoding='utf-8') as f:
for line in data:
json_record = json.dumps(line, ensure_ascii=False)
f.write(json_record + '\n')
if __name__ == "__main__":
result = []
corpus = parse(xmlfiles[0])
(counter.extend([line["THREAD_SEQUENCE"] for line in corpus]))
output_path = "train.jsonl"
write_to_file(corpus,output_path)
corpus = parse(xmlfiles[1])
(counter.extend([line["THREAD_SEQUENCE"] for line in corpus]))
output_path = "train.jsonl"
write_to_file(corpus,output_path,"a")
print(Counter(counter))
corpus = parse(xmlfiles[2])
(counter.extend([line["THREAD_SEQUENCE"] for line in corpus]))
output_path = "val.jsonl"
write_to_file(corpus,output_path)
print(Counter(counter))
