Build a quote tagger for motivational content

Creating a quote tagger for motivational content involves building a system that can automatically identify and tag motivational quotes in text. This can be useful for organizing, filtering, or enhancing motivational articles, social media posts, or apps.

Here’s a detailed plan and example implementation of a motivational quote tagger using Python with NLP techniques:

Approach Overview

1. Data Collection: Gather a dataset of motivational quotes and non-motivational sentences.

2. Preprocessing: Clean and prepare text data.

3. Feature Extraction: Convert text to numerical features (e.g., TF-IDF vectors).

4. Model Training: Use a classification model (e.g., Logistic Regression, SVM) to tag motivational content.

5. Tagging: Predict whether new sentences are motivational quotes.

6. Optional: Use keyword-based heuristics or embeddings for improved accuracy.

Step-by-step Implementation

1. Sample Data Preparation

You can start with a small dataset of motivational quotes and non-motivational sentences.

python
CopyEdit
motivational_quotes = [
    "Believe in yourself and all that you are.",
    "Success is not final, failure is not fatal: It is the courage to continue that counts.",
    "The only way to do great work is to love what you do.",
    "Don’t watch the clock; do what it does. Keep going.",
    "Your limitation—it’s only your imagination."
]

non_motivational_sentences = [
    "The weather today is cloudy with a chance of rain.",
    "I need to buy groceries after work.",
    "She went to the market to buy fresh vegetables.",
    "The cat is sleeping on the couch.",
    "It is important to save money for emergencies."
]

2. Preprocessing & Labeling

Create labels for training data:

python
CopyEdit
texts = motivational_quotes + non_motivational_sentences
labels = [1] * len(motivational_quotes) + [0] * len(non_motivational_sentences)

3. Feature Extraction & Model Training

python
CopyEdit
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LogisticRegression
from sklearn.pipeline import make_pipeline

# Split data
X_train, X_test, y_train, y_test = train_test_split(texts, labels, test_size=0.2, random_state=42)

# Build pipeline with TF-IDF and Logistic Regression
model = make_pipeline(TfidfVectorizer(), LogisticRegression())

# Train
model.fit(X_train, y_train)

# Evaluate
print(f"Accuracy on test set: {model.score(X_test, y_test):.2f}")

4. Tagging New Content

python
CopyEdit
def tag_quotes(texts):
    predictions = model.predict(texts)
    return ["Motivational" if pred == 1 else "Non-Motivational" for pred in predictions]

# Example usage
new_texts = [
    "Hard work beats talent when talent doesn't work hard.",
    "I went to the park yesterday.",
    "Every day is a second chance."
]

tags = tag_quotes(new_texts)
for text, tag in zip(new_texts, tags):
    print(f"'{text}' => {tag}")

Enhancements for Real-World Use

• Expand Dataset: Use large datasets of quotes and general sentences for better accuracy.

• Use Word Embeddings: Use pre-trained embeddings (e.g., BERT, GloVe) for semantic understanding.

• Keyword-Based Heuristics: Detect motivational keywords (like success, believe, dream, courage).

• Sentence Length & Punctuation: Motivational quotes often have distinct punctuation and length patterns.

• Contextual Models: Fine-tune transformer models (like BERT) for better performance.

This solution can be integrated into web apps, content management systems, or social media tools to automatically tag motivational quotes in a body of text. Would you like me to provide a more advanced example using deep learning or a ready-to-use API approach?