Fake/real news classification [tensorflow (keras)]

Giskard is an open-source framework for testing all ML models, from LLMs to tabular models. Don’t hesitate to give the project a star on GitHub ⭐️ if you find it useful!

In this notebook, you’ll learn how to create comprehensive test suites for your model in a few lines of code, thanks to Giskard’s open-source Python library.

Use-case:

• Binary classification of news being fake or real, based on their text.

• Reference notebook

• Dataset

Outline:

• Detect vulnerabilities automatically with Giskard’s scan

• Automatically generate & curate a comprehensive test suite to test your model beyond accuracy-related metrics

Install dependencies

Make sure to install the giskard

%pip install giskard --upgrade

Import libraries

import os
import string
import tarfile
from pathlib import Path
from typing import Tuple, Callable
from urllib.request import urlretrieve

import numpy as np
import pandas as pd
from keras.layers import Dense, Embedding, LSTM
from keras.models import Sequential
from keras.optimizers import Adam
from keras.preprocessing.text import Tokenizer
from keras.utils import pad_sequences
from nltk.corpus import stopwords
from sklearn.metrics import accuracy_score
from sklearn.model_selection import train_test_split

from giskard import Dataset, Model, scan, testing

Notebook-level settings

os.environ["TF_CPP_MIN_LOG_LEVEL"] = "1"

Define constants

# Constants.
MAX_TOKENS = 20000
MAX_SEQUENCE_LENGTH = 100
N_ROWS = 1000

STOPWORDS = stopwords.words("english")

TEXT_COLUMN_NAME = "text"
TARGET_COLUMN_NAME = "isFake"

RANDOM_SEED = 0

# Paths.
DATA_URL = "https://giskard-library-test-datasets.s3.eu-north-1.amazonaws.com/fake_real_news_dataset-{}"
DATA_PATH = Path.home() / ".giskard" / "fake_real_news_dataset"

Dataset preparation

Load and preprocess data

def fetch_demo_data(url: str, file: Path) -> None:
    """Helper to fetch data from the FTP server."""
    if not file.parent.exists():
        file.parent.mkdir(parents=True, exist_ok=True)

    if not file.exists():
        print(f"Downloading data from {url}")
        urlretrieve(url, file)

    print(f"Data was loaded!")


def fetch_dataset() -> None:
    """Gradually fetch all necessary files from the FTP server."""
    files_to_fetch = ("Fake.csv.tar.gz", "True.csv.tar.gz", "glove_100d.txt.tar.gz")
    for file_name in files_to_fetch:
        fetch_demo_data(DATA_URL.format(file_name), DATA_PATH / file_name)


def load_data(**kwargs) -> pd.DataFrame:
    """Load data."""
    real_df = pd.read_csv(DATA_PATH / "True.csv.tar.gz", **kwargs)
    fake_df = pd.read_csv(DATA_PATH / "Fake.csv.tar.gz", **kwargs)

    # Create target column.
    real_df[TARGET_COLUMN_NAME] = 0
    fake_df[TARGET_COLUMN_NAME] = 1

    # Combine dfs.
    full_df = pd.concat([real_df, fake_df])
    full_df.drop(columns=["subject", "date"], inplace=True)
    return full_df


fetch_dataset()
news_df = load_data(nrows=N_ROWS)

Train-test split

X_train, X_test, Y_train, Y_test = train_test_split(
    news_df[["title", TEXT_COLUMN_NAME]], news_df[TARGET_COLUMN_NAME], random_state=RANDOM_SEED
)

Wrap dataset with Giskard

To prepare for the vulnerability scan, make sure to wrap your dataset using Giskard’s Dataset class. More details here.

raw_data = pd.concat([X_test, Y_test], axis=1)
giskard_dataset = Dataset(
    df=raw_data,
    # A pandas.DataFrame that contains the raw data (before all the pre-processing steps) and the actual ground truth variable (target).
    target=TARGET_COLUMN_NAME,  # Ground truth variable.
    name="fake_and_real_news",  # Optional.
)

Model building

Define preprocessing steps

def prepare_text(df: pd.DataFrame) -> np.ndarray:
    """Perform text-data cleaning: punctuation and stop words removal."""
    # Merge text data into single column.
    df[TEXT_COLUMN_NAME] = df[TEXT_COLUMN_NAME] + " " + df.title
    df.drop(columns=["title"], inplace=True)

    # Remove punctuation.
    df[TEXT_COLUMN_NAME] = df[TEXT_COLUMN_NAME].apply(
        lambda text: text.translate(str.maketrans("", "", string.punctuation))
    )

    # Remove stop words.
    df[TEXT_COLUMN_NAME] = df[TEXT_COLUMN_NAME].apply(
        lambda sentence: " ".join([_word for _word in sentence.split() if _word.lower() not in STOPWORDS])
    )

    return df[TEXT_COLUMN_NAME]


X_train_prepared = prepare_text(X_train)
X_test_prepared = prepare_text(X_test)


def init_tokenizer() -> Tuple[Callable, Tokenizer]:
    """Initialize tokenization function with the Tokenizer in it's outer-scope."""
    tokenizer = Tokenizer(num_words=MAX_TOKENS)
    tokenizer.fit_on_texts(X_train_prepared)

    def tokenization_closure(df: pd.DataFrame) -> pd.DataFrame:
        tokenized = tokenizer.texts_to_sequences(df)
        return pad_sequences(tokenized, maxlen=MAX_SEQUENCE_LENGTH)

    return tokenization_closure, tokenizer


tokenize, text_tokenizer = init_tokenizer()
X_train_tokens = tokenize(X_train_prepared)
X_test_tokens = tokenize(X_test_prepared)

Create embeddings matrix

def parse_line(word: str, *arr: list) -> Tuple[str, np.ndarray]:
    """Parse line from the file with embeddings.
    The first value of the line is the word and the rest values are related glove embedding: (<word>, 0.66, 0.23, ...)."""
    return word, np.asarray(arr, dtype="float32")


def init_embeddings_matrix(embeddings_dict: dict) -> np.ndarray:
    """Init a matrix, where each row is an embedding vector."""
    num_embeddings = min(MAX_TOKENS, len(text_tokenizer.word_index))
    stacked_embeddings = np.stack(list(embeddings_dict.values()))
    embeddings_mean, embeddings_std, embeddings_dimension = (
        stacked_embeddings.mean(),
        stacked_embeddings.std(),
        stacked_embeddings.shape[1],
    )
    embeddings_matrix = np.random.normal(embeddings_mean, embeddings_std, (num_embeddings, embeddings_dimension))
    return embeddings_matrix


def get_embeddings_matrix() -> np.ndarray:
    """Create matrix, where each row is an embedding of a specific word."""
    # Load glove embeddings.
    embeddings_dict = dict(
        parse_line(*line.rstrip().rsplit(" "))
        for line in tarfile.open(DATA_PATH / "glove_100d.txt.tar.gz", "r:gz")
        .extractfile("fake_real_news_dataset-glove_100d.txt")
        .read()
        .decode()
    )

    # Create embeddings matrix with glove word vectors.
    embeddings_matrix = init_embeddings_matrix(embeddings_dict)
    for word, idx in text_tokenizer.word_index.items():
        if idx >= MAX_TOKENS:
            continue

        embedding_vector = embeddings_dict.get(word, None)

        if embedding_vector is not None:
            embeddings_matrix[idx] = embedding_vector

    return embeddings_matrix


embed_matrix = get_embeddings_matrix()

Build estimator

def init_model() -> Sequential:
    """Initialize new TF model."""
    # Define model container.
    model = Sequential()

    # Non-trainable embedding layer.
    model.add(
        Embedding(MAX_TOKENS, output_dim=100, weights=[embed_matrix], input_length=MAX_SEQUENCE_LENGTH, trainable=False)
    )

    # LSTM stage.
    model.add(LSTM(units=32, return_sequences=True, recurrent_dropout=0.25, dropout=0.25))
    model.add(LSTM(units=16, recurrent_dropout=0.1, dropout=0.1))

    # Dense stage.
    model.add(Dense(units=16, activation="relu"))
    model.add(Dense(units=1, activation="sigmoid"))

    # Build model.
    model.compile(optimizer=Adam(learning_rate=0.01), loss="binary_crossentropy", metrics=["accuracy"])
    return model


# Fit model.
n_epochs = 5
batch_size = 256

classifier = init_model()
_ = classifier.fit(
    X_train_tokens, Y_train, batch_size=batch_size, validation_data=(X_test_tokens, Y_test), epochs=n_epochs
)

train_metric = classifier.evaluate(X_train_tokens, Y_train, verbose=0)[1]
test_metric = classifier.evaluate(X_test_tokens, Y_test, verbose=0)[1]

print(f"Train accuracy: {train_metric: .4f}")
print(f"Test accuracy: {test_metric: .4f}")

Wrap model with Giskard

To prepare for the vulnerability scan, make sure to wrap your model using Giskard’s Model class. You can choose to either wrap the prediction function (preferred option) or the model object. More details here.

def prediction_function(df: pd.DataFrame) -> np.ndarray:
    """Define a prediction function for giskard.Model."""
    tokens = tokenize(prepare_text(df))
    return classifier.predict(tokens, verbose=0)


giskard_model = Model(
    model=prediction_function,
    # A prediction function that encapsulates all the data pre-processing steps and that could be executed with the dataset used by the scan.
    model_type="classification",  # Either regression, classification or text_generation.
    name="fake_real_news_classification",  # Optional.
    feature_names=["title", TEXT_COLUMN_NAME],  # Default: all columns of your dataset.
    classification_labels=[0, 1],  # Their order MUST be identical to the prediction_function's output order.
    # classification_threshold=0.5  # Default: 0.5
)

# Validate wrapped model.
Y_test_pred_wrapper = giskard_model.predict(giskard_dataset).prediction
wrapped_test_metric = accuracy_score(Y_test, Y_test_pred_wrapper)
print(f"Wrapped test accuracy: {wrapped_test_metric: .4f}")

Detect vulnerabilities in your model

Scan your model for vulnerabilities with Giskard

Giskard’s scan allows you to detect vulnerabilities in your model automatically. These include performance biases, unrobustness, data leakage, stochasticity, underconfidence, ethical issues, and more. For detailed information about the scan feature, please refer to our scan documentation.

results = scan(giskard_model, giskard_dataset)

display(results)

Generate comprehensive test suites automatically for your model

Generate test suites from the scan

The objects produced by the scan can be used as fixtures to generate a test suite that integrate all detected vulnerabilities. Test suites allow you to evaluate and validate your model’s performance, ensuring that it behaves as expected on a set of predefined test cases, and to identify any regressions or issues that might arise during development or updates.

test_suite = results.generate_test_suite("My first test suite")
test_suite.run()

2024-05-29 12:04:25,717 pid:57157 MainThread giskard.datasets.base INFO     Casting dataframe columns from {'title': 'object', 'text': 'object'} to {'title': 'object', 'text': 'object'}
2024-05-29 12:04:25,720 pid:57157 MainThread giskard.utils.logging_utils INFO     Predicted dataset with shape (30, 3) executed in 0:00:00.013458
Executed 'Recall on data slice “`text_length(text)` >= 2570.500 AND `text_length(text)` < 2729.500”' with arguments {'model': <giskard.models.function.PredictionFunctionModel object at 0x30d36ec80>, 'dataset': <giskard.datasets.base.Dataset object at 0x31dc1b820>, 'slicing_function': <giskard.slicing.text_slicer.MetadataSliceFunction object at 0x366023850>, 'threshold': 0.9384146341463414}:
               Test failed
               Metric: 0.91


2024-05-29 12:04:25,756 pid:57157 MainThread giskard.datasets.base INFO     Casting dataframe columns from {'title': 'object', 'text': 'object'} to {'title': 'object', 'text': 'object'}
2024-05-29 12:04:25,757 pid:57157 MainThread giskard.utils.logging_utils INFO     Predicted dataset with shape (42, 3) executed in 0:00:00.006029
Executed 'Recall on data slice “`text` contains "texas"”' with arguments {'model': <giskard.models.function.PredictionFunctionModel object at 0x30d36ec80>, 'dataset': <giskard.datasets.base.Dataset object at 0x31dc1b820>, 'slicing_function': <giskard.slicing.slice.QueryBasedSliceFunction object at 0x316d33850>, 'threshold': 0.9384146341463414}:
               Test failed
               Metric: 0.91


2024-05-29 12:04:25,781 pid:57157 MainThread giskard.datasets.base INFO     Casting dataframe columns from {'title': 'object', 'text': 'object'} to {'title': 'object', 'text': 'object'}
2024-05-29 12:04:25,782 pid:57157 MainThread giskard.utils.logging_utils INFO     Predicted dataset with shape (38, 3) executed in 0:00:00.005618
Executed 'Recall on data slice “`text` contains "october"”' with arguments {'model': <giskard.models.function.PredictionFunctionModel object at 0x30d36ec80>, 'dataset': <giskard.datasets.base.Dataset object at 0x31dc1b820>, 'slicing_function': <giskard.slicing.slice.QueryBasedSliceFunction object at 0x316caf160>, 'threshold': 0.9384146341463414}:
               Test failed
               Metric: 0.91


2024-05-29 12:04:25,808 pid:57157 MainThread giskard.datasets.base INFO     Casting dataframe columns from {'title': 'object', 'text': 'object'} to {'title': 'object', 'text': 'object'}
2024-05-29 12:04:25,808 pid:57157 MainThread giskard.utils.logging_utils INFO     Predicted dataset with shape (27, 3) executed in 0:00:00.005583
Executed 'Accuracy on data slice “`text` contains "guilty"”' with arguments {'model': <giskard.models.function.PredictionFunctionModel object at 0x30d36ec80>, 'dataset': <giskard.datasets.base.Dataset object at 0x31dc1b820>, 'slicing_function': <giskard.slicing.slice.QueryBasedSliceFunction object at 0x344927e80>, 'threshold': 0.9424}:
               Test failed
               Metric: 0.93


2024-05-29 12:04:25,832 pid:57157 MainThread giskard.datasets.base INFO     Casting dataframe columns from {'title': 'object', 'text': 'object'} to {'title': 'object', 'text': 'object'}
2024-05-29 12:04:25,833 pid:57157 MainThread giskard.utils.logging_utils INFO     Predicted dataset with shape (61, 3) executed in 0:00:00.006272
Executed 'Recall on data slice “`text` contains "investigation"”' with arguments {'model': <giskard.models.function.PredictionFunctionModel object at 0x30d36ec80>, 'dataset': <giskard.datasets.base.Dataset object at 0x31dc1b820>, 'slicing_function': <giskard.slicing.slice.QueryBasedSliceFunction object at 0x31837ea70>, 'threshold': 0.9384146341463414}:
               Test failed
               Metric: 0.93


2024-05-29 12:04:25,844 pid:57157 MainThread giskard.datasets.base INFO     Casting dataframe columns from {'title': 'object', 'text': 'object'} to {'title': 'object', 'text': 'object'}
2024-05-29 12:04:25,845 pid:57157 MainThread giskard.utils.logging_utils INFO     Predicted dataset with shape (33, 3) executed in 0:00:00.006265
Executed 'Recall on data slice “`text_length(title)` >= 92.500 AND `text_length(title)` < 97.500”' with arguments {'model': <giskard.models.function.PredictionFunctionModel object at 0x30d36ec80>, 'dataset': <giskard.datasets.base.Dataset object at 0x31dc1b820>, 'slicing_function': <giskard.slicing.text_slicer.MetadataSliceFunction object at 0x34709ebf0>, 'threshold': 0.9384146341463414}:
               Test failed
               Metric: 0.93


2024-05-29 12:04:25,847 pid:57157 MainThread giskard.core.suite INFO     Executed test suite 'My first test suite'
2024-05-29 12:04:25,847 pid:57157 MainThread giskard.core.suite INFO     result: failed
2024-05-29 12:04:25,848 pid:57157 MainThread giskard.core.suite INFO     Recall on data slice “`text_length(text)` >= 2570.500 AND `text_length(text)` < 2729.500” ({'model': <giskard.models.function.PredictionFunctionModel object at 0x30d36ec80>, 'dataset': <giskard.datasets.base.Dataset object at 0x31dc1b820>, 'slicing_function': <giskard.slicing.text_slicer.MetadataSliceFunction object at 0x366023850>, 'threshold': 0.9384146341463414}): {failed, metric=0.9090909090909091}
2024-05-29 12:04:25,848 pid:57157 MainThread giskard.core.suite INFO     Recall on data slice “`text` contains "texas"” ({'model': <giskard.models.function.PredictionFunctionModel object at 0x30d36ec80>, 'dataset': <giskard.datasets.base.Dataset object at 0x31dc1b820>, 'slicing_function': <giskard.slicing.slice.QueryBasedSliceFunction object at 0x316d33850>, 'threshold': 0.9384146341463414}): {failed, metric=0.9090909090909091}
2024-05-29 12:04:25,849 pid:57157 MainThread giskard.core.suite INFO     Recall on data slice “`text` contains "october"” ({'model': <giskard.models.function.PredictionFunctionModel object at 0x30d36ec80>, 'dataset': <giskard.datasets.base.Dataset object at 0x31dc1b820>, 'slicing_function': <giskard.slicing.slice.QueryBasedSliceFunction object at 0x316caf160>, 'threshold': 0.9384146341463414}): {failed, metric=0.9130434782608695}
2024-05-29 12:04:25,849 pid:57157 MainThread giskard.core.suite INFO     Accuracy on data slice “`text` contains "guilty"” ({'model': <giskard.models.function.PredictionFunctionModel object at 0x30d36ec80>, 'dataset': <giskard.datasets.base.Dataset object at 0x31dc1b820>, 'slicing_function': <giskard.slicing.slice.QueryBasedSliceFunction object at 0x344927e80>, 'threshold': 0.9424}): {failed, metric=0.9259259259259259}
2024-05-29 12:04:25,850 pid:57157 MainThread giskard.core.suite INFO     Recall on data slice “`text` contains "investigation"” ({'model': <giskard.models.function.PredictionFunctionModel object at 0x30d36ec80>, 'dataset': <giskard.datasets.base.Dataset object at 0x31dc1b820>, 'slicing_function': <giskard.slicing.slice.QueryBasedSliceFunction object at 0x31837ea70>, 'threshold': 0.9384146341463414}): {failed, metric=0.9333333333333333}
2024-05-29 12:04:25,850 pid:57157 MainThread giskard.core.suite INFO     Recall on data slice “`text_length(title)` >= 92.500 AND `text_length(title)` < 97.500” ({'model': <giskard.models.function.PredictionFunctionModel object at 0x30d36ec80>, 'dataset': <giskard.datasets.base.Dataset object at 0x31dc1b820>, 'slicing_function': <giskard.slicing.text_slicer.MetadataSliceFunction object at 0x34709ebf0>, 'threshold': 0.9384146341463414}): {failed, metric=0.9333333333333333}

Test suite failed.

Test Recall on data slice “`text_length(text)` >= 2570.500 AND `text_length(text)` < 2729.500”

Measured Metric = 0.90909 Failed

model fake_real_news_classification

dataset fake_and_real_news

slicing_function `text_length(text)` >= 2570.500 AND `text_length(text)` < 2729.500

threshold 0.9384146341463414

Test Recall on data slice “`text` contains "texas"”

Measured Metric = 0.90909 Failed

model fake_real_news_classification

dataset fake_and_real_news

slicing_function `text` contains "texas"

threshold 0.9384146341463414

Test Recall on data slice “`text` contains "october"”

Measured Metric = 0.91304 Failed

model fake_real_news_classification

dataset fake_and_real_news

slicing_function `text` contains "october"

threshold 0.9384146341463414

Test Accuracy on data slice “`text` contains "guilty"”

Measured Metric = 0.92593 Failed

model fake_real_news_classification

dataset fake_and_real_news

slicing_function `text` contains "guilty"

threshold 0.9424

Test Recall on data slice “`text` contains "investigation"”

Measured Metric = 0.93333 Failed

model fake_real_news_classification

dataset fake_and_real_news

slicing_function `text` contains "investigation"

threshold 0.9384146341463414

Test Recall on data slice “`text_length(title)` >= 92.500 AND `text_length(title)` < 97.500”

Measured Metric = 0.93333 Failed

model fake_real_news_classification

dataset fake_and_real_news

slicing_function `text_length(title)` >= 92.500 AND `text_length(title)` < 97.500

threshold 0.9384146341463414

Customize your suite by loading objects from the Giskard catalog

The Giskard open source catalog will enable to load:

• Tests such as metamorphic, performance, prediction & data drift, statistical tests, etc

• Slicing functions such as detectors of toxicity, hate, emotion, etc

• Transformation functions such as generators of typos, paraphrase, style tune, etc

To create custom tests, refer to this page.

For demo purposes, we will load a simple unit test (test_f1) that checks if the test F1 score is above the given threshold. For more examples of tests and functions, refer to the Giskard catalog.

test_suite.add_test(testing.test_f1(model=giskard_model, dataset=giskard_dataset, threshold=0.7)).run()