def romance_similarity(movieA, movieB): score = 0 score += shared_tropes_weight(tropeA, tropeB) * 3 score -= abs(agency_indexA - agency_indexB) * 1.5 score += if family_interference_level_close() * 2 score += shared_song_mood_bonus() return score Example: Liked "Yeh Jawaani Hai Deewani" → Recommend "Zindagi Na Milegi Dobara" (friends-to-lovers + travel backdrop) and "Tamasha" (identity + romance). When user clicks on a film:
"pair": "Aditya + Geet", "arc": "Depressed businessman meets talkative small-town girl → separates → reunites by choice", "power_dynamic": "Female leads emotional growth", "tropes": ["runaway bride", "train journey", "cold male melts"], "score": "intensity": 6.2, "agency": 9, "family_control": 2, "songs_worth_watching": ["Mauja Hi Mauja", "Nagada Nagada"] , "verdict": "Healthy, uplifting. Geet remains herself throughout."
This feature turns Bollywood romance from passive watching into an – perfect for a streaming platform, fan community, or film studies tool.
| Love Language | Bollywood Cue | |---------------|----------------| | Words of Affirmation | Shayari, letters, "Maine pyar kiya" | | Acts of Service | Fighting goons, fixing court case | | Gifts | Mangalsutra, dupatta, sketch | | Quality Time | Rickshaw rides, chai at tapri | | Physical Touch | Accidental hand touch in rain |
🎯 Core Purpose Analyze, categorize, and visualize the dynamics of romantic relationships in Bollywood films—helping users discover movies based on relationship type, emotional arc, and cultural tropes. 1. Data Model – Relationship Taxonomy Define a JSON schema for each romantic storyline:
📖 SUMMARY OF ARC: Muslim boy meets Hindu girl at university. Families object. Secret meetings at temple/mosque. Third-act court scene. Finally, “Tum kisi ki roko naa…” acceptance.
💔 ROMANCE TYPE: Forbidden Love (Interfaith) 📈 INTENSITY: 8.5/10 👑 AGENCY: She convinces family (6/10) 🌧️ GRAND GESTURE: Climax – runs away from wedding mandap
🎵 DEFINING SONG: “Tum Hi Ho” – longing, separation, solo male.