def load_review_data():
df = pd.read_csv("data/Reviews.csv")
# preview
print(df.head())
print(df.tail())
vocab = []
X = np.zeros((len(df), 2), dtype=object)
for i in tqdm.tqdm(range(len(df)), "Cleaning X"):
target = df['Text'].loc[i]
X[i, 0] = clean_text(target)
X[i, 1] = df['Score'].loc[i]
for word in X[i, 0].split():
vocab.append(word)
# vocab = set(vocab)
vocab = Counter(vocab)
# delete words that occur less than 10 times
vocab = {k: v for k, v in vocab.items() if v >= N}
# word to integer encoder dict
vocab2int = {word: i for i, word in enumerate(vocab, start=1)}
# pickle int2vocab for testing
print("Pickling vocab2int...")
pickle.dump(vocab2int, open("data/vocab2int.pickle", "wb"))
# encoded reviews
for i in tqdm.tqdm(range(X.shape[0]), "Tokenizing words"):
X[i, 0] = tokenize_words(X[i, 0], vocab2int)
lengths = [len(row) for row in X[:, 0]]
print("min_length:", min(lengths))
print("max_length:", max(lengths))
X_train, X_test, y_train, y_test = train_test_split(X[:, 0],
X[:, 1],
test_size=test_size,
shuffle=True,
random_state=19)
return X_train, X_test, y_train, y_test, vocab
def load_emotion_data():
df = pd.read_csv("data/text_emotion.csv")
additional_anger = read_text_file("data/anger.txt")
# additional_happy = read_text_file("data/happy.txt")
# additional_sadness = read_text_file("data/sadness.txt")
# additional_files = [additional_anger, additional_happy, additional_sadness]
additional_files = [additional_anger]
print(df.head())
print(df.tail())
# calculate number of data samples
# for our additional files
n_samples = 0
for file in additional_files:
n_samples += len(file)
print("samples:", n_samples)
vocab = []
X = np.zeros((len(df) + n_samples, 2), dtype=object)
for i in tqdm.tqdm(range(len(df)), "Cleaning data"):
target = df['content'].loc[i]
try:
emotion = categories_reversed[df['sentiment'].loc[i]]
except KeyError:
continue
X[i, 0] = clean_text(target)
X[i, 1] = emotion
for word in X[i, 0].split():
vocab.append(word)
k = i + 1
for file in tqdm.tqdm(additional_files, "Cleaning additional data"):
for i, (text, emotion) in enumerate(file, start=k):
X[i, 0] = clean_text(text)
if emotion == "joy":
emotion = "happiness"
X[i, 1] = categories_reversed[emotion]
for word in X[i, 0].split():
vocab.append(word)
k = i
# remove zero lines
# X = X[np.any(X != 0, axis=1)]
vocab = Counter(vocab)
# delete words that occur less than 10 times
vocab = {k: v for k, v in vocab.items() if v >= N}
# word to integer encoder dict
vocab2int = {word: i for i, word in enumerate(vocab, start=1)}
print("Pickling vocab2int...")
pickle.dump(vocab2int, open("data/vocab2int.pickle", "wb"))
# encoded reviews
for i in tqdm.tqdm(range(X.shape[0]), "Tokenizing words"):
X[i, 0] = tokenize_words(str(X[i, 0]), vocab2int)
lengths = [len(row) for row in X[:, 0]]
print("min_length:", min(lengths))
print("max_length:", max(lengths))
X_train, X_test, y_train, y_test = train_test_split(X[:, 0],
X[:, 1],
test_size=test_size,
shuffle=True,
random_state=7)
return X_train, X_test, y_train, y_test, vocab
audio = r.listen(source)
print("Loading vocab2int")
vocab2int = pickle.load(open("Mood:Emotion Code/data/vocab2int.pickle", "rb"))
model = get_model_emotions(len(vocab2int),
sequence_length=sequence_length,
embedding_size=embedding_size)
model.load_weights("results/model_v1_0.59_0.76.h5")
if __name__ == "__main__":
import argparse
# parser = argparse.ArgumentParser(description="Emotion classifier using text")
# parser.add_argument("text", type=str, help="The text you want to analyze")
# args = parser.parse_args()
text = tokenize_words(clean_text(r.recognize_google(audio)), vocab2int)
x = pad_sequences([text], maxlen=sequence_length)
prediction = model.predict_classes(x)[0]
probs = model.predict(x)[0]
# print("hi:",index)
print("Question asked: ", Textlist[index])
print("You said: " + r.recognize_google(audio))
print("Probs:")
for i, category in categories.items():
print(f"{category.capitalize()}: {probs[i]*100:.2f}%")
print("The most dominant emotion:", categories[prediction])
# audio = r.listen(source)
print("Loading vocab2int")
vocab2int = pickle.load(open("data/vocab2int.pickle", "rb"))
model = get_model_emotions(len(vocab2int),
sequence_length=sequence_length,
embedding_size=embedding_size)
model.load_weights("results/model_v1_0.59_0.76.h5")
if __name__ == "__main__":
import argparse
parser = argparse.ArgumentParser(
description="Emotion classifier using text")
parser.add_argument("text", type=str, help="The text you want to analyze")
args = parser.parse_args()
text = tokenize_words(clean_text(args.text), vocab2int)
x = pad_sequences([text], maxlen=sequence_length)
prediction = model.predict_classes(x)[0]
probs = model.predict(x)[0]
# print("hi:",index)
# print("Question asked: ",Textlist[index])
# print("You said: " + r.recognize_google(audio))
print("Probs:")
for i, category in categories.items():
print(f"{category.capitalize()}: {probs[i]*100:.2f}%")
print("The most dominant emotion:", categories[prediction])
'CPU': 1,
'GPU': 0
})
from model import get_model_5stars
from utils import clean_text, tokenize_words
from config import embedding_size, sequence_length
from keras.preprocessing.sequence import pad_sequences
import pickle
vocab2int = pickle.load(open("data/vocab2int.pickle", "rb"))
model = get_model_5stars(len(vocab2int),
sequence_length=sequence_length,
embedding_size=embedding_size)
model.load_weights("results/model_V20_0.38_0.80.h5")
if __name__ == "__main__":
import argparse
parser = argparse.ArgumentParser(description="Food Review evaluator")
parser.add_argument("review",
type=str,
help="The review of the product in text")
args = parser.parse_args()
review = tokenize_words(clean_text(args.review), vocab2int)
x = pad_sequences([review], maxlen=sequence_length)
print(f"{model.predict(x)[0][0]:.2f}/5")