NLP LAB ASSIGNMENT -05
NAME : Raj Kumar Reddy
REG.NO: 22BCE9821
SLOT : L16+17
1.Implement a text classification application?
CODE :
#Implement a text classification application using
import tensorflow as tf
from tensorflow.keras.preprocessing.text import Tokenizer
from tensorflow.keras.preprocessing.sequence import pad_sequences
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Embedding, LSTM, Dense
import numpy as np
sentences = [
"This is a positive sentence.",
"This is another negative sentence."
]
labels = [1, 0]
# Tokenize the text data
tokenizer = Tokenizer(num_words=5000)
tokenizer.fit_on_texts(sentences)
sequences = tokenizer.texts_to_sequences(sentences)
# Pad sequences to a fixed length
max_length = max([len(seq) for seq in sequences])
padded_sequences = pad_sequences(sequences, maxlen=max_length,
padding='post')
# Build the RNN model
model = Sequential()
model.add(Embedding(5000, 128, input_length=max_length))
model.add(LSTM(128))
model.add(Dense(1, activation='sigmoid'))
# Compile the model
model.compile(optimizer='adam', loss='binary_crossentropy',
metrics=['accuracy'])
# Train the model
# Convert padded_sequences and labels to NumPy arrays explicitly
padded_sequences = np.array(padded_sequences)
labels = np.array(labels)
model.fit(padded_sequences, labels, epochs=10)
# Example prediction
new_sentence = "This is a great and positive sentence."
new_sequence = tokenizer.texts_to_sequences([new_sentence])
new_padded_sequence = pad_sequences(new_sequence, maxlen=max_length,
padding='post')
prediction = model.predict(new_padded_sequence)
if prediction > 0.5:
print("Positive sentiment")
else:
print("Negative sentiment")
OUTPUT :
Epoch 1/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 3s 3s/step - accuracy:
0.5000 - loss: 0.6946
Epoch 2/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 129ms/step - accuracy:
1.0000 - loss: 0.6886
Epoch 3/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 59ms/step - accuracy:
1.0000 - loss: 0.6827
Epoch 4/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 59ms/step - accuracy:
1.0000 - loss: 0.6764
Epoch 5/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 58ms/step - accuracy:
1.0000 - loss: 0.6697
Epoch 6/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 63ms/step - accuracy:
1.0000 - loss: 0.6624
Epoch 7/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 65ms/step - accuracy:
1.0000 - loss: 0.6541
Epoch 8/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 139ms/step - accuracy:
1.0000 - loss: 0.6449
Epoch 9/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 60ms/step - accuracy:
1.0000 - loss: 0.6343
Epoch 10/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 60ms/step - accuracy:
1.0000 - loss: 0.6223
2.Build a sentiment analyzer using LSTM.
CODE :
# Build a sentiment analyzer using LSTM taking different example
import tensorflow as tf
from tensorflow.keras.preprocessing.text import Tokenizer
from tensorflow.keras.preprocessing.sequence import pad_sequences
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Embedding, LSTM, Dense
import numpy as np
# Sample data (replace with your own dataset)
sentences = [
"I love this product, it's amazing!",
"This is the worst experience ever.",
"The movie was fantastic and entertaining.",
"I'm really disappointed with the service.",
"The food was delicious and the service was excellent.",
"I would never recommend this place to anyone."
]
labels = [1, 0, 1, 0, 1, 0] # 1 for positive, 0 for negative
# Tokenize the text data
tokenizer = Tokenizer(num_words=5000)
tokenizer.fit_on_texts(sentences)
sequences = tokenizer.texts_to_sequences(sentences)
# Pad sequences to a fixed length
max_length = max([len(seq) for seq in sequences])
padded_sequences = pad_sequences(sequences, maxlen=max_length,
padding='post')
# Build the RNN model
model = Sequential()
model.add(Embedding(5000, 128, input_length=max_length))
model.add(LSTM(128))
model.add(Dense(1, activation='sigmoid'))
# Compile the model
model.compile(optimizer='adam', loss='binary_crossentropy',
metrics=['accuracy'])
# Train the model
# Convert padded_sequences and labels to NumPy arrays explicitly
padded_sequences = np.array(padded_sequences)
labels = np.array(labels)
model.fit(padded_sequences, labels, epochs=10)
# Example prediction
new_sentence = "The product is okay, nothing special."
new_sequence = tokenizer.texts_to_sequences([new_sentence])
new_padded_sequence = pad_sequences(new_sequence, maxlen=max_length,
padding='post')
prediction = model.predict(new_padded_sequence)
if prediction > 0.5:
print("Positive sentiment")
else:
print("Negative sentiment")
OUTPUT :
Epoch 1/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 3s 3s/step - accuracy:
0.1667 - loss: 0.6941
Epoch 2/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 68ms/step - accuracy:
0.6667 - loss: 0.6891
Epoch 3/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 66ms/step - accuracy:
0.6667 - loss: 0.6841
Epoch 4/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 63ms/step - accuracy:
0.8333 - loss: 0.6785
Epoch 5/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 62ms/step - accuracy:
0.8333 - loss: 0.6718
Epoch 6/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 69ms/step - accuracy:
0.8333 - loss: 0.6637
Epoch 7/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 60ms/step - accuracy:
1.0000 - loss: 0.6535
Epoch 8/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 60ms/step - accuracy:
1.0000 - loss: 0.6405
Epoch 9/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 141ms/step - accuracy:
1.0000 - loss: 0.6241
Epoch 10/10
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 64ms/step - accuracy:
1.0000 - loss: 0.6032
