Skip to content

yoginiaishwaryaa/script-match

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

17 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🎬 ScriptMatch

Automated Movie Recommendation via AI Transcription

ScriptMatch is a fully serverless, Azure-native platform that recommends movies based on what's actually said in them — not just their genre or manual tags. It uses AI to transcribe movie audio, extract meaningful keywords from the dialogue, and match films that share similar narrative content.


The Problem It Solves

Traditional movie recommendation systems rely on broad genre labels and manually assigned tags. This approach is slow, inconsistent, and misses the richest source of information a movie has — its actual dialogue. ScriptMatch fixes this by automatically analyzing what characters say, extracting key vocabulary, and using that to surface genuinely similar films.


How It Works

1. Video Ingestion

A movie is uploaded to Azure Blob Storage. This automatically triggers Azure Event Grid, which kicks off the processing pipeline via Azure Logic Apps — no manual intervention needed.

2. AI Transcription & Keyword Extraction

Azure AI Video Indexer processes the video and extracts:

  • Full transcripts
  • Keywords and topics
  • Sentiment signals

Each movie ends up with a rich semantic metadata profile built entirely from its content.

3. Keyword Indexing

The extracted keywords and metadata are stored in Azure AI Search, making every movie fully searchable and comparable by content rather than category.

4. Movie Matching & Recommendation

When a user watches a movie, their taste profile is updated in Azure Cosmos DB — tracking the genres and keywords they engage with. ScriptMatch then computes similarity between movies (and users) based on shared keyword patterns, surfacing recommendations that are genuinely content-driven.


Azure Services Used

Service Role
Azure Blob Storage Stores uploaded video files
Azure Event Grid Detects uploads and triggers workflows
Azure Logic Apps Orchestrates the processing pipeline
Azure AI Video Indexer Transcribes audio and extracts keywords
Azure AI Search Indexes and queries movie metadata
Azure Functions Runs all backend logic (fully serverless)
Azure Web App Hosting web applications, APIs, and mobile backends

Tech Stack

  • Backend — Node.js with Express
  • Frontend — HTML, CSS, Vanilla JavaScript
  • Cloud — Microsoft Azure (fully serverless, no containers or VMs)

👥 Team

Name GitHub
Reenu L K ReenuLK
Krishna Deepak krisndeep
Yogini Aishwaryaa P T S yoginiaishwaryaa

About

An AI-powered, serverless platform that recommends movies by analyzing and matching their dialogue using automated transcription and keyword extraction.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors